Hostname: page-component-586b7cd67f-t7fkt Total loading time: 0 Render date: 2024-11-22T13:04:50.284Z Has data issue: false hasContentIssue false

Processed meat: the real villain?

Published online by Cambridge University Press:  01 December 2015

Sabine Rohrmann*
Affiliation:
Division of Chronic Disease Epidemiology, Epidemiology, Biostatistics and Prevention Institute, University of Zurich, Zurich, Switzerland
Jakob Linseisen
Affiliation:
Institute of Epidemiology II, Helmholtz Zentrum München, Neuherberg, Germany
*
*Corresponding author: S. Rohrmann, email [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Meat is a food rich in protein, minerals such as iron and zinc as well as a variety of vitamins, in particular B vitamins. However, the content of cholesterol and saturated fat is higher than in some other food groups. Processed meat is defined as products usually made of red meat that are cured, salted or smoked (e.g. ham or bacon) in order to improve the durability of the food and/or to improve colour and taste, and often contain a high amount of minced fatty tissue (e.g. sausages). Hence, high consumption of processed foods may lead to an increased intake of saturated fats, cholesterol, salt, nitrite, haem iron, polycyclic aromatic hydrocarbons, and, depending upon the chosen food preparation method, also heterocyclic amines. Several large cohort studies have shown that a high consumption of processed (red) meat is related to increased overall and cause-specific mortality. A meta-analysis of nine cohort studies observed a higher mortality among high consumers of processed red meat (relative risk (RR) = 1·23; 95 % CI 1·17, 1·28, top v. bottom consumption category), but not unprocessed red meat (RR = 1·10; 95 % CI 0·98, 1·22). Similar associations were reported in a second meta-analysis. All studies argue that plausible mechanisms are available linking processed meat consumption and risk of chronic diseases such as CVD, diabetes mellitus or some types of cancer. However, the results of meta-analyses do show some degree of heterogeneity between studies, and it has to be taken into account that individuals with low red or processed meat consumption tend to have a healthier lifestyle in general. Hence, substantial residual confounding cannot be excluded. Information from other types of studies in man is needed to support a causal role of processed meat in the aetiology of chronic diseases, e.g. studies using the Mendelian randomisation approach.

Type
Conference on ‘The future of animal products in the human diet: health and environmental concerns’
Copyright
Copyright © The Authors 2015 

Meat is an integral part of human diet in many cultures and in recent years, meat consumption has increased considerably in most parts of the world. According to the Food and Agriculture Organization, world per capita meat consumption was just over 30 kg per person per year in 1980; in 2005, it was 41 kg per person per year. As incomes rise, more meat is consumed( 1 ). Meat consumption contributes to intake of a number of vitamins and minerals such as vitamin B, vitamin A, zinc and iron. It is also an important source of protein providing essential amino acids; however, meat is also rich in saturated fat and cholesterol. The total fat content of meat varies considerably with average values (median, interquartile range) of 9·3 (5·1–15) g/100 g in beef and 12·4 (4–16·2) g/100 g in pork; the same is true for SFA with values of 3·8 (2·5–6·1) g /100 g in beef and 3·5 (1·4–5·5) g/100 g in pork( Reference Payne, Scarborough and Rayner 2 ). Total fat and saturated fat concentrations are distinctly higher in many types of processed meat, with extreme values of up to 90 g/100 g total fat and 25 g/100 g saturated fat in fatty bacon. Calculations of the contribution of meat and processed meat to the total daily energy intake or the total daily fat intake underpin the important role of meats in present dietary practice( Reference Linseisen, Kesse and Slimani 3 ).

Meat has already been consumed for thousands of years. Prior to the availability of adequate storage such as refrigerators or deep-freezers that can preserve fresh meat for a longer period, preservation by drying, salting, curing or smoking were the only means to be able to provide meat in times when no fresh meat was available. For example, salting and sun-drying was used in ancient Egypt( Reference Pearson and Gillett 4 ). In addition to improving the meat's durability, meat processing is also used to preserve the food's colour and taste.

Salting, i.e. adding NaCl to meat, increases its durability by decreasing the water content of meat and inhibiting micro-organisms( Reference Toldra and Flores 5 ). Curing, that is adding salt enriched with nitrates and nitrites to meat products for preservation purposes, leads to the formation of N-nitroso compounds (NOC) and increases the originally low salt (NaCl) content of fresh meat. Similar to the developments in meat smoking, developments in manufacturing practice, e.g. addition of ascorbic acid, decreased the amount of nitrate/nitrite added to processed meat products in most European countries during past decades( Reference Honikel 6 ).

In addition to this exogenous exposure, there is endogenous nitrate and nitrite generation from inducible and endogenous NO synthases, resulting in NOC production. NOC arise from the reaction of nitrite and secondary or tertiary amines in the intestine from N-nitrosation of amines, which are produced in the colon by bacterial decarboxylation of amino acids( Reference Bingham, Hughes and Cross 7 ). Additionally, haem iron from red but not white meat substantially increases the endogenous production of NOC( Reference Bingham, Hughes and Cross 7 ).

Smoking of meat inactivates enzymes and micro-organisms and influences its taste. As a downside, however, smoke contains polycyclic aromatic hydrocarbons (PAH), which are formed by pyrolytic processes at high smoking temperatures (400–1000°C). The type of wood used, the temperature, the use of smoke flavour additives and whether direct or indirect, hot or cold smoking methods determine the amount of PAH that is produced during this process( Reference Simko 8 ). In the past couple of years, even decades, with improvements in managing the smoking process, the amount of PAH that is produced has decreased considerably( Reference Behsnilian, Butz and Greiner 9 ). In addition to PAH production, smoking also increases the concentration of NOC in food( Reference Lijinsky 10 , Reference Haorah, Zhou and Wang 11 ).

In this review, we evaluate the current evidence on the association of processed meat consumption with mortality and the incidence of cancer, CVD and diabetes. In addition, we discuss effects of different components in red and processed meat and their possible role in the aetiology of these chronic diseases.

Total mortality

Several prospective studies evaluated the association between meat intake and mortality( Reference Kahn, Phillips and Snowdon 12 Reference Rohrmann, Overvad and Bueno-de-Mesquita 24 ), some of which compared meat consumers with vegetarians( Reference Fraser 14 , Reference Appleby, Key and Thorogood 16 , Reference Chang-Claude, Hermann and Eilber 18 , Reference Key, Appleby and Spencer 20 ). The results of these studies mostly pointed in the direction of a positive association in particular of processed meat, consumption and all-cause mortality. Three meta-analyses have been published in the past 2 years using basically the same studies and coming up with similar results( Reference Abete, Romaguera and Vieira 25 Reference Wang, Lin and Ouyang 27 ). In all of them, processed meat consumption was significantly associated with a moderately increased all-cause mortality (Table 1), but the consumption of unprocessed red meat was not.

Table 1. Association between red and processed meat consumption and all-cause and cause-specific mortality: results of meta-analyses

Several of the earlier mentioned cohorts estimated the contribution of high processed meat consumption to total mortality in terms of attributable or preventable fractions. The results of these estimates were quite heterogeneous: In the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort, we estimated that a reduction of processed meat consumption to <20 g/d could reduce the total mortality by 3·3 % (95 % CI 1·5, 5·0 %). In contrast, estimates from US cohorts were much higher. In the American Association of Retired Persons cohort, the preventable fraction was estimated to be 20 % if women decreased their processed meat consumption to less than 1·6 g per 4184 kJ/d (1·6 g per 1000 kcal/d)( Reference Sinha, Cross and Graubard 19 ) and in two other US cohorts, the preventable fraction was estimated to be 9·3 % in the Health Professionals Follow-up Study and 7·6 % in the Nurses’ Health Study if the participants lowered their red meat (processed and unprocessed) consumption to less than 0·5 servings daily. The difference between the US studies and the EPIC result is likely due to the stronger risk estimates observed in the US cohorts compared with EPIC, but may also be explained by higher meat consumption in the US than in Europe as well as differences in meat preparation and cooking.

An analysis of EPIC-Oxford participants observed that vegetarians as well as non-vegetarians with a health-conscious lifestyle had a statistically significantly lower mortality compared with the British general population( Reference Key, Appleby and Spencer 20 ). Similarly, in a study in Germany, both vegetarians and health-conscious non-vegetarians had a statistically significantly lower overall mortality compared with the general population( Reference Chang-Claude, Hermann and Eilber 18 ). The implication of these results is that perhaps the decreased mortality in vegetarians, compared with the general population, is in large part due to a healthier lifestyle, such as having lower body fat, being more physically active, and not being a smoker. However, in the large US and European cohorts that did indeed report an increased risk for early mortality among individuals with a high red and processed meat consumption, this increase in risk was largely independent of smoking, obesity and other potential confounders( Reference Sinha, Cross and Graubard 19 , Reference Pan, Sun and Bernstein 23 , Reference Rohrmann, Overvad and Bueno-de-Mesquita 24 ).

Cancer

The cancer entity that is best studied in relation to meat consumption is colorectal cancer. Numerous case–control and cohort studies have evaluated the question whether red and/or processed meat consumption is associated with risk of this disease. Many of the studies, case–control as well as cohort studies, did indeed observe positive associations between red meat consumption and colorectal cancer risk. In a summary evaluation of the studies published thus far, the World Cancer Research Fund in 2007 came to the conclusion that high consumption of unprocessed red meat and of processed meat were convincingly associated with the risk of colorectal cancer( 28 ). This was confirmed in the 2011 updated report. Per 50 g increase in daily processed meat consumption, the relative risk (RR) increased by 18 % (95 % CI 1·10, 1·28); per 100 g increase in daily unprocessed red meat consumption, the RR increased by 17 % (95 % CI 1·05, 1·31)( 29 ). Underlining this, the International Agency for Research on Cancer recently declared that there is sufficient evidence in human beings for the carcinogenicity of the consumption of processed meat( Reference Bouvard, Loomis and Guyton 30 ).

In addition to colorectal cancer, high consumption of red and processed meat might be linked to several other cancer entities such as oesophagus, lung, pancreas and endometrium (red meat) as well as oesophagus, lung and stomach (processed meat)( 28 ).

Based on these findings, World Cancer Research Fund recommended in their conclusions that red meat intake should be limited to <500 g/week and very little, if any, of this should be processed meat.

Several mechanisms were proposed and examined to explain an increased risk of certain types of cancer with increased meat, particularly processed meat consumption (Table 2).

Table 2. Potential mechanisms connecting meat consumption and risk of chronic diseases

T2D, type-2 diabetes mellitus; PAH, polycyclic aromatic hydrocarbons; HCA, heterocyclic aromatic amines; AGE, advanced glycation end products; TMAO, trimethylamine-N-oxide; NOC, N-nitroso compounds.

Nitrites or nitrates added to meat for preservation could increase exogenous exposure to nitrosamines, NOC and their precursors. Dietary intake of NOC was associated with cancer risk, in particular gastrointestinal cancer. For example, in the EPIC-Norfolk study, N-nitrosodimethylamine intake was associated with increased risk of gastrointestinal cancers (hazard ratio (HR) = 1·13; 95 % CI 1·00, 1·28), specifically of rectal cancer (HR = 1·46; 95 % CI 1·16, 1·84 per 1-sd increase)( Reference Loh, Jakszyn and Luben 31 ).

High intake of salt and consumption of salted and salty foods is considered a probable risk factor for gastric cancer( 28 , Reference D'Elia, Rossi and Ippolito 32 ). Some traditional diets include substantial amounts of salt-preserved foods, including salted meat, fish or vegetables and salted foods such as bacon, sausages and ham, which contain from 3 to 5 g salt/ 100 g( 28 ). High salt intake can damage the lining of the stomach, increase endogenous NOC formation, synergistically interact with gastric carcinogens, and increase the colonisation by Helicobacter pylori ( 28 , Reference D'Elia, Rossi and Ippolito 32 ).

Haem iron in red meat may lead to oxidative stress, which, in turn, might increase peroxidation of lipids, lead to protein modification and DNA damage( Reference Bastide, Chenni and Audebert 33 ). Haem iron also increases endogenous NOC formation  because haem in red meat can readily become nitrosylated and act as a nitrosating agent. Based on the results of the EPIC cohort, endogenous NOC formation may account for the association between red and processed meat consumption and gastric cancer risk( Reference Jakszyn, Bingham and Pera 34 ), and in a Shanghai cohort, higher  endogenous NOC formation was associated with increased risk of colorectal cancer( Reference Dellavalle, Xiao and Yang 35 ).

Heterocyclic aromatic amines (HCA) and PAH are considered carcinogenic( 36 ) and have long been hypothesised to contribute to cancer risk( Reference Abid, Cross and Sinha 37 ). Several epidemiology studies did indeed observe positive associations between dietary intake of HCA and PAH with risk of different types of cancer, in particular colorectal adenomas and colorectal cancer( Reference Rohrmann, Hermann and Linseisen 38 Reference Ferrucci, Sinha and Graubard 40 ). However, the study results are still rather heterogeneous, which might partly be due to crude dietary assessment methods but also differences across study populations with respect to genetic background of phases I and  II enzymes responsible for HCA and PAH metabolism( Reference Abid, Cross and Sinha 37 , Reference Turesky and Le Marchand 41 ).

A long-standing theory suggests that high saturated fat intake increases the risk of a variety of cancer types. In the USA, American Association of Retired Persons as well as in the EPIC cohort, high intake of saturated fats was associated with an increased risk of postmenopausal breast cancer, in particular among women who never used postmenopausal hormones( Reference Thiebaut, Kipnis and Chang 42 , Reference Sieri, Krogh and Ferrari 43 ). These results support earlier findings of a pooled analysis of eight cohorts, in which high saturated fat intake tended to be associated with increased breast cancer risk (pooled RR = 1·09; 95 % CI 1·00, 1·19 for an increment of 5 % of energy from saturated fat)( Reference Smith-Warner, Spiegelman and Adami 44 ). However, for most cancer types, results are rather heterogeneous( 28 ) or point towards no associations, such as for prostate cancer( Reference Xu, Han and Zeng 45 ).

A new hypothesis has been proposed by zur Hausen( Reference zur Hausen 46 ). Based on the observation that the worldwide distribution of colorectal cancer correlates with rates of beef consumption, he proposed that a specific beef factor, suspected to be one or more thermoresistant potentially oncogenic bovine viruses could contaminate beef preparations leading subsequently to latent infections in the colorectal tract. Preceding, concomitant or subsequent exposure to chemical carcinogens arising during cooking procedures can then result in increased risk for colorectal cancer synergistic with these infections( Reference zur Hausen 46 ). So far, however, no epidemiological studies have addressed this hypothesis.

CVD

As for colorectal cancer several studies have examined the association between red and processed meat consumption and the risk of CVD, in particular, myocardial infarction, or more broadly CHD, but also stroke. In contrast to colorectal cancer, for which both unprocessed and processed red meat appear to be important, the most recent meta-analysis showed a significantly positive association of processed meat intake with CHD (RR = 1·42; 95 % CI 1·07, 1·89 per 50 g increase in daily consumption), but not with consumption of unprocessed red meat (RR = 1·00; 95 % CI 0·82, 1·23)( Reference Micha, Michas and Mozaffarian 47 ).

Less frequently studied is the association of meat consumption with the risk of stroke. In a meta-analysis that included six cohort studies( Reference Kaluza, Wolk and Larsson 48 ), the risk of total stroke increased by 11 % (95 % CI 1·03, 1·20) for each serving per d increase in fresh red meat and by 13 % (95 % CI 1·03, 1·24) for processed meat. The authors did not detect large heterogeneity among studies (P > 0·16). Both fresh and processed red meat were related to an increased risk of ischaemic stroke (RR = 1·13; 95 % CI 1·00, 1·27 and RR = 1·15; 95 % CI 1·06, 1·24, respectively). However, neither meat type was related to the risk of haemorrhagic stroke (fresh red meat RR = 1·08, 95 % CI 0·84, 1·39; processed meat RR = 1·16, 95 % CI 0·92, 1·46).

Several hypotheses have been formulated to explain the associations of processed meat with the risk of CVD. Adding salt (NaCl) to red meat for conservation purposes increases the naturally low sodium content of red meat. In their meta-analysis, Micha et al. stated that processed meats contain about 400 % more sodium and 50 % more nitrates per g( Reference Micha, Michas and Mozaffarian 47 ), although this depends strongly on the type of meat and the methods used( Reference Linseisen, Rohrmann and Norat 49 ). A high salt intake is thought to be associated with hypertension and, consequently, an increased risk of CVD( Reference He, Li and Macgregor 50 , Reference Aburto, Ziolkovska and Hooper 51 ), although it is currently still unclear which amounts of salt intake do affect blood pressure and whether only certain subgroups of the population would particularly benefit from decreasing their salt intake( Reference O'Donnell, Mente and Yusuf 52 ). Processed meats such as sausages, salami and bacon have a higher content of SFA and cholesterol than fresh red meat; the latter is often consumed after removing the visible fat tissue, whereas the proportion of fat in sausages often reaches 50 % of weight or even more. Although numerous studies have been conducted on the association between fat intake and risk of CHD, the association appears to be yet rather unclear. A 2010 meta-analysis came to the conclusion that both high saturated fat and cholesterol intake are related to the risk of CHD( Reference Mozaffarian, Micha and Wallace 53 ), but a more recent one did not find a convincing association between dietary intake of saturated fats and coronary disease( Reference Chowdhury, Warnakula and Kunutsor 54 ). However, their effects on blood lipoproteins are well described and the latter are established causal factors in the aetiology of CVD. Nitrates and their byproducts (e.g. peroxynitrite) experimentally promote endothelial dysfunction, atherosclerosis and insulin resistance( Reference Micha, Michas and Mozaffarian 47 ).

Some other potential mechanisms do not only apply to processed meat, but to red meat in general. Firstly, haem iron in red meat may lead to oxidative stress, which, in turn, might increase peroxidation of lipids, lead to protein modification and DNA damage. Results of some studies suggested that high body iron stores (e.g. serum ferritin) could be determinants of levels of systemic oxidative DNA damage( Reference Hori, Mizoue and Kasai 55 , Reference Nakano, Kawanishi and Kamohara 56 ) and some, but not all epidemiological studies have shown associations between body iron stores and risk of myocardial infarction( Reference Iqbal, Mehboobali and Tareen 57 ). Secondly, higher intake of red meat is related to higher intake of arachidonic acid, which leads to higher plasma concentration( Reference Astorg, Bertrais and Laporte 58 ). This may cause changes in fatty acid concentration and pattern of fatty acids of platelet membranes, and eicosanoids produced from arachidonic acid promote inflammatory and prothrombotic activities. However, dietary intake of arachidonic acid does not appear to be related to the risk of stroke( Reference Sakai, Kakutani and Tokuda 59 ) and the association of dietary or circulating arachidonic acid with CHD is yet unclear( Reference Mozaffarian, Ascherio and Hu 60 Reference Reinders, van Ballegooijen and Visser 62 ). More recently, US studies observed that CVD patients with higher concentrations of trimethylamine-N-oxide (TMAO) have a higher risk for major adverse cardiovascular events such as death, myocardial infarction or stroke than patients with low TMAO concentrations( Reference Tang, Wang and Levison 63 ). Intestinal bacteria metabolise the precursor of TMAO, trimethylamine, from carnitine, phosphatidylcholine (lecithin) and choline. After absorption, in a second step, trimethylamine is oxidised to TMAO in the liver( Reference Koeth, Wang and Levison 64 , Reference Wang, Klipfell and Bennett 65 ). These trimethylamine precursors, carnitine, lecithin (phosphatidylcholine) and choline, are abundant in red meat and liver, but also fish, milk, cheese and eggs( Reference Hamlin, Pauly and Melnyk 66 ). So far, however, it is unclear if and how dietary intake of red meat or any other food affects circulating TMAO concentration in healthy individuals ( Reference Rohrmann, Linseisen and Allenspach 67 ).

Type-2 diabetes mellitus

A meta-analysis conducted by Micha et al. concluded that both unprocessed red meat and processed meat consumption were associated with an increased risk of type-2 diabetes. Per 50 g increase in daily processed meat consumption, the risk increased by 51 % (95 % CI 1·25, 1·81; eight cohorts), whereas the association was less strong with unprocessed red meat intake (RR = 1·19; 95 % CI 1·04, 1·37; nine cohorts; per 100 g intake)( Reference Micha, Michas and Mozaffarian 47 ). In an evaluation of the EPIC cohort, using a case–cohort design, red meat (HR: 1·08; 95 % CI 1·03, 1·13), processed meat (HR: 1·12; 95 % CI 1·05, 1·19) and meat iron intake (HR: 1·03; 95 % CI 0·99, 1·07) were associated with incident type-2 diabetes( Reference Bendinelli, Palli and Masala 68 ).

In an analysis of the α-Tocopherol β-Carotin Cancer Prevention cohort, the positive association between processed meat consumption and diabetes risk was explained more by dietary intake of sodium than by intake of SFA, protein, cholesterol, haem iron, magnesium and nitrate, and these results were not modified by obesity( Reference Mannisto, Kontto and Kataja-Tuomola 69 ). However, it is yet unclear how high sodium intake might contribute to type-2 diabetes aetiology, although salt restriction in diabetes patients might be beneficial for blood pressure control( Reference Suckling, He and Macgregor 70 ).

Based on the results of animal models, some authors hypothesised that chronic exposure to nitrosamine compounds could contribute to the pathogenesis of type-2 diabetes( Reference Tong, Neusner and Longato 71 ). Nitrosamines activated during metabolism may generate reactive oxygen species, which, in turn, can increase oxidative stress, DNA damage, lipid peroxidation and protein adduct formation. Oxidative stress and DNA damage lead to activation of pro-inflammatory cytokines and insulin resistance( Reference de la Monte, Neusner and Chu 72 ).

As for cancer and CVD, haem iron appears to be an important factor in the association between red and processed meat consumption and risk of diabetes, which is supported by the EPIC-Interact results( Reference Bendinelli, Palli and Masala 68 ). A large meta-analysis reported strong associations of serum ferritin concentration and clinically elevated transferrin saturation with an increased risk of type-2 diabetes( Reference Orban, Schwab and Thorand 73 ). These associations were even seen after adjusting for inflammatory factors. It is thought that higher body iron stores might impair insulin sensitivity and increase the risk of diabetes by promoting oxidative stress causing tissue damage( Reference Rajpathak, Crandall and Wylie-Rosett 74 ).

High amounts of advanced glycation end products are found in animal products high in protein and fat, such as meats and cheeses. In addition, higher concentrations were seen in (industrially) processed foods from animal products such as frankfurters, bacon and powdered egg whites, compared with the unprocessed forms( Reference Peppa, Goldberg and Cai 75 ). It is well-known that high circulating advanced glycation end products  levels are associated with adverse outcomes in diabetes patients( Reference Malmstedt, Karvestedt and Swedenborg 76 Reference Hanssen, Beulens and van Dieren 78 ), but so far no epidemiological study has evaluated whether dietary advanced glycation end products intake or circulating levels are associated with incident type-2 diabetes.

Inflammation appears to be involved in mediating the association between red meat consumption and diabetes (as well as CVD). In the EPIC-Potsdam study, a cohort with about 25 000 participants, a high consumption of red meat was associated with higher circulating levels of γ-glytamyl transferase and high-sensitivity C-reactive protein( Reference Montonen, Boeing and Fritsche 79 ). Similarly, higher red meat consumption was associated with unfavourable plasma concentrations of inflammatory and glucose metabolic markers in diabetes-free participants of the Nurses’ Health Study( Reference Ley, Sun and Willett 80 ). Interestingly, BMI accounted for a significant proportion of the observed associations with these biomarkers, except for ferritin (see the next section). The authors concluded from their analysis that the substitution of red meat with other protein food would be related to a healthier biomarker profile of inflammatory and glucose metabolism.

Methodological considerations

Interaction with other foods and nutrients

In EPIC, Norat et al. observed a strong positive association between red and processed meat intake and risk of colorectal cancer( Reference Norat, Bingham and Ferrari 81 ). However, depending on other dietary habits, i.e. fish consumption and fibre intake, the associations were different. For example, the HR in study participants with high intake of red and processed meat was 1·09 (95 % CI 0·83, 1·42) for the group with high intake of fibre (>26 g/d in women, >28 g/d in men), but 1·50 (95 % CI 1·15, 1·97) for the group with low intake of fibre (<17 g/d) compared with participants who had low intake of red and processed meat and high intake of fibre (P interaction = 0·06). Similar interaction was observed by fish consumption. Also, in a study on HCA intake and colorectal adenoma risk, we observed a stronger association between the intake of 2-amino-1-methyl-6-phenylimidazo(4,5-b)pyridine and adenoma risk in individuals with a flavonol intake below the median intake in the cohort, the RR progressively increased with higher 2-amino-1-methyl-6-phenylimidazo(4,5-b)pyridine intake (RR = 1·76; 95 % CI 1·17, 2·65; P trend 0·01; top v. bottom quartile). However, no statistically significant associations were observed for participants with a high flavonol intake (RR = 1·24, 95 % CI 0·85, 1·80; P trend 0·14; top v. bottom quartile)( Reference Rohrmann, Hermann and Linseisen 38 ). This observation is consistent with results from experimental studies in which interactive effects of phases I and II enzymes on the risk of HCA-associated cancers have been described( Reference Hodek, Trefil and Stiborova 82 ).

A third example is a potential interaction of nitrite intake from diet with intake of polyphenols on risk of gastric cancer. In a Mexcian study, a high intake of total nitrite as well as nitrate and nitrite from animal sources doubled the risk of gastric cancer (OR = 1·92; 95 % CI 1·23, 3·02, top v. bottom tertile)( Reference Hernandez-Ramirez, Galvan-Portillo and Ward 83 ). OR about 2-fold were observed among individuals with both low intake of cinnamic acids, secoisolariciresinol or coumestrol and high intake of animal-derived nitrate or nitrite compared with high intake of the polyphenols and low animal nitrate or nitrite intake. The results of this study suggest that polyphenols may reduce gastric cancer risk through inhibition of endogenous nitrosation( Reference Hernandez-Ramirez, Galvan-Portillo and Ward 83 ). In a similar way, vitamin C intake modified the association between N-nitrosodimethylamine intake and risk of gastrointestinal cancers in the EPIC-Norfolk cohort( Reference Loh, Jakszyn and Luben 31 ).

Residual confounding

Problems in epidemiological studies are factors that act as confounders, i.e. are related to both the exposure and the outcome. Incomplete adjustment for such confounders result in residual confounding; this applies to factors that have not been assessed at all or to factors that have not been assessed in sufficient detail or precision. In many studies, incomplete adjustment for active (and passive) smoking may pose a problem. In the EPIC analysis on the association between meat consumption and mortality, we observed heterogeneity according to smoking (P interaction 0·019), with significant associations between processed meat intake and all-cause mortality only in former and current smokers but no significant associations in never smokers( Reference Rohrmann, Overvad and Bueno-de-Mesquita 24 ).

Heterogeneity between studies

As described earlier, several meta-analyses have been conducted using studies from a variety of settings that differ in time, place and type of dietary assessment. All these factors may contribute to heterogeneity in study results. For example, in their meta-analysis on meat consumption and all-cause mortality, Larsson et al. pointed out that, although most studies observed positive associations, formal tests for heterogeneity were statistically significant( Reference Larsson and Orsini 26 ). This heterogeneity might be due to different meat consumption habits and, thus, differences in the range of red and processed meat consumption in a population (e.g. higher in the USA than in East Asia), to differences in foods that contribute to meat consumption categories (e.g. different types of processed meats consumed in different populations) and to the length of follow-up( Reference Larsson and Orsini 26 ). Another factor that contributes to heterogeneity in study results is differences in adjustment variables as shown in a meta-analysis of meat consumption and colorectal cancer risk( Reference Hernandez-Ramirez, Galvan-Portillo and Ward 84 ).

Conclusions

Processed meat, which is mostly processed red meat, but might also include white meat, is associated with increased all-cause mortality and also with increased risk of some types of cancer (such as colorectal and gastric cancer), CVD and type-2 diabetes. Although most epidemiological studies point towards such an association, the strength of the association appears to be unclear. For example, the associations between processed meat consumption and all-cause mortality appear to be much stronger in the USA( Reference Sinha, Cross and Graubard 19 , Reference Pan, Sun and Bernstein 23 ) than among European( Reference Rohrmann, Overvad and Bueno-de-Mesquita 24 ) studies. The reasons for this discrepancy are still unclear.

Factors that are associated with total meat and in particular with processed meat consumption, and can act as confounders need to be addressed carefully in epidemiological studies as lifestyle differs between individuals with high and low processed meat consumption. Information from other types of studies in human subjects are needed to support a causal role of processed meat in the aetiology of chronic diseases. Using the Bradford Hill criteria simply based on epidemiological studies is not sufficient because it does not preclude misinterpretation due to confounding or bias( Reference Mente, de Koning and Shannon 85 ). However, trials that randomise individuals into a low consumption v. control group are difficult to conduct, in particular if the outcome is a ‘hard’ endpoint, such as cancer, myocardial infarction or diabetes. Using intermediate endpoints, such as changes in blood lipids, concentration of advanced glycation end products or DNA adducts is difficult, too, because the link between these markers and incident disease or mortality is not unique, such that high cholesterol concentrations might be linked to CHD but also some types of cancer. Other approaches, such as studies using the Mendelian randomisation approach, may help to establish causal associations between processed meat consumption and risk of chronic diseases.

If, however, after careful evaluation of existing studies, it will turn out that the processed meat consumption is indeed causally associated with chronic diseases, it needs to be addressed which factors are responsible for these associations and how the risk might be reduced.

Financial Support

None.

Conflicts of Interest

None.

Authorship

S. R. drafted the manuscript. J. L. reviewed and revised the manuscript.

References

1. Food and Agriculture Organization of the United Nations (FAO) (2006) Lifestock's Long Shadow – Environmental Issues and Options. Rome: FAO.Google Scholar
2. Payne, CL, Scarborough, P, Rayner, M et al. (2015) Are edible insects more or less ‘healthy’ than commonly consumed meats? A comparison using two nutrient profiling models developed to combat over- and undernutrition. Eur J Clin Nutr. [Epublication ahead of print]Google Scholar
3. Linseisen, J, Kesse, E, Slimani, N et al. (2002) Meat consumption in the European prospective investigation into cancer and nutrition (EPIC) cohorts: results from 24-hour dietary recalls. Public Health Nutr 5, 12431258.CrossRefGoogle ScholarPubMed
4. Pearson, A & Gillett, T (editors) (1996) Processed Meats, 3rd ed. Philadelphia: Springer Science & Business Media.Google Scholar
5. Toldra, F & Flores, M (1998) The role of muscle proteases and lipases in flavor development during the processing of dry-cured ham. Crit Rev Food Sci Nutr 38, 331352.CrossRefGoogle ScholarPubMed
6. Honikel, KO (editor) (1995) The Use of Additives in Meat Products Throughout Europe. Necessity, Customs, Legislation. Utrecht: ECCEAMST.Google Scholar
7. Bingham, SA, Hughes, R & Cross, AJ (2002) Effect of white versus red meat on endogenous N-nitrosation in the human colon and further evidence of a dose response. J Nutr 132, 3522S3525S.Google Scholar
8. Simko, P (2002) Determination of polycyclic aromatic hydrocarbons in smoked meat products and smoke flavouring food additives. J Chromatogr B Analyt Technol Biomed Life Sci 770, 318.Google Scholar
9. Behsnilian, D, Butz, P, Greiner, R et al. (2014) Process-induced undesirable compounds: chances of non-thermal approaches. Meat Sci 98, 392403.Google Scholar
10. Lijinsky, W (1999) N-nitroso compounds in the diet. Mutat Res 443, 129138.Google Scholar
11. Haorah, J, Zhou, L, Wang, X et al. (2001) Determination of total N-nitroso compounds and their precursors in frankfurters, fresh meat, dried salted fish, sauces, tobacco, and tobacco smoke particulates. J Agric Food Chem 49, 60686078.Google Scholar
12. Kahn, HA, Phillips, RL, Snowdon, DA et al. (1984) Association between reported diet and all-cause mortality. Twenty-one-year follow-up on 27,530 adult Seventh-Day Adventists. Am J Epidemiol 119, 775787.Google Scholar
13. Whiteman, D, Muir, J, Jones, L et al. (1999) Dietary questions as determinants of mortality: the OXCHECK experience. Public Health Nutr 2, 477487.Google Scholar
14. Fraser, GE (1999) Associations between diet and cancer, ischemic heart disease, and all-cause mortality in non-Hispanic white California Seventh-day Adventists. Am J Clin Nutr 70, 532S538S.Google Scholar
15. Fortes, C, Forastiere, F, Farchi, S et al. (2000) Diet and overall survival in a cohort of very elderly people. Epidemiology 11, 440445.Google Scholar
16. Appleby, PN, Key, TJ, Thorogood, M et al. (2002) Mortality in British vegetarians. Public Health Nutr 5, 2936.Google Scholar
17. Kelemen, LE, Kushi, LH, Jacobs, DR Jr et al. (2005) Associations of dietary protein with disease and mortality in a prospective study of postmenopausal women. Am J Epidemiol 161, 239249.Google Scholar
18. Chang-Claude, J, Hermann, S, Eilber, U et al. (2005) Lifestyle determinants and mortality in German vegetarians and health-conscious persons: results of a 21-year follow-up. Cancer Epidemiol Biomarkers Prev 14, 963968.Google Scholar
19. Sinha, R, Cross, AJ, Graubard, BI et al. (2009) Meat intake and mortality: a prospective study of over half a million people. Arch Intern Med 169, 562571.Google Scholar
20. Key, TJ, Appleby, PN, Spencer, EA et al. (2009) Mortality in British vegetarians: results from the European Prospective Investigation into Cancer and Nutrition (EPIC-Oxford). Am J Clin Nutr 89, 1613S1619S.Google Scholar
21. Trichopoulou, A, Bamia, C & Trichopoulos, D (2009) Anatomy of health effects of Mediterranean diet: Greek EPIC prospective cohort study. BMJ 338, b2337.CrossRefGoogle ScholarPubMed
22. Nagao, M, Iso, H, Yamagishi, K et al. (2012) Meat consumption in relation to mortality from cardiovascular disease among Japanese men and women. Eur J Clin Nutr 66, 687693.Google Scholar
23. Pan, A, Sun, Q, Bernstein, AM et al. (2012) Red meat consumption and mortality: results from 2 prospective cohort studies. Arch Intern Med 172, 555563.Google Scholar
24. Rohrmann, S, Overvad, K, Bueno-de-Mesquita, HB et al. (2013) Meat consumption and mortality – results from the European Prospective Investigation into Cancer and Nutrition. BMC Med 11, 63.CrossRefGoogle ScholarPubMed
25. Abete, I, Romaguera, D, Vieira, AR et al. (2014) Association between total, processed, red and white meat consumption and all-cause, CVD and IHD mortality: a meta-analysis of cohort studies. Br J Nutr 112, 762775.Google Scholar
26. Larsson, SC & Orsini, N (2014) Red meat and processed meat consumption and all-cause mortality: a meta-analysis. Am J Epidemiol 179, 282289.Google Scholar
27. Wang, X, Lin, X, Ouyang, YY et al. (2015) Red and processed meat consumption and mortality: dose-response meta-analysis of prospective cohort studies. Public Health Nutr [Epublication ahead of print].Google Scholar
28. American Institute for Cancer Research/World Cancer Research Fund (2007) Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective [AICR, editor]. Washington, DC: AICR.Google Scholar
29. World Cancer Research Fund/American Institute for Cancer Research (2011) Food, Nutrition, Physical Activity, and the Prevention of Colorectal Cancer. Washington, DC: AICR.Google Scholar
30. Bouvard, Vr, Loomis, D, Guyton, KZ et al. (2015) Carcinogenicity of consumption of red and processed meat. Lancet Oncol [Epublication ahead of print].Google Scholar
31. Loh, YH, Jakszyn, P, Luben, RN et al. (2011) N-Nitroso compounds and cancer incidence: the European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk study. Am J Clin Nutr 93, 10531061.Google Scholar
32. D'Elia, L, Rossi, G, Ippolito, R et al. (2012) Habitual salt intake and risk of gastric cancer: a meta-analysis of prospective studies. Clin Nutr 31, 489498.Google Scholar
33. Bastide, NM, Chenni, F, Audebert, M et al. (2015) A central role for heme iron in colon carcinogenesis associated with red meat intake. Cancer Res 75, 870879.Google Scholar
34. Jakszyn, P, Bingham, S, Pera, G et al. (2006) Endogenous versus exogenous exposure to N-nitroso compounds and gastric cancer risk in the European Prospective Investigation into Cancer and Nutrition (EPIC-EURGAST) study. Carcinogenesis 27, 14971501.Google Scholar
35. Dellavalle, CT, Xiao, Q, Yang, G et al. (2014) Dietary nitrate and nitrite intake and risk of colorectal cancer in the Shanghai Women's Health Study. Int J Cancer 134, 29172926.Google Scholar
36. International Agency for Research on Cancer (1993) Heterocyclic Aromatic Amines. Some Naturally Occurring Substances: Food Items and Constituents, Heterocyclic Aromatic Amines and Mycotoxins, pp. 165242. Lyon: IARC.Google Scholar
37. Abid, Z, Cross, AJ & Sinha, R (2014) Meat, dairy, and cancer. Am J Clin Nutr 100, Suppl. 1, 386S393S.CrossRefGoogle ScholarPubMed
38. Rohrmann, S, Hermann, S & Linseisen, J (2009) Heterocyclic aromatic amine intake increases colorectal adenoma risk: findings from a prospective European cohort study. Am J Clin Nutr 89, 14181424.Google Scholar
39. Cross, AJ, Ferrucci, LM, Risch, A et al. (2010) A large prospective study of meat consumption and colorectal cancer risk: an investigation of potential mechanisms underlying this association. Cancer Res 70, 24062414.CrossRefGoogle ScholarPubMed
40. Ferrucci, LM, Sinha, R, Graubard, BI et al. (2009) Dietary meat intake in relation to colorectal adenoma in asymptomatic women. Am J Gastroenterol 104, 12311240.Google Scholar
41. Turesky, RJ & Le Marchand, L (2011) Metabolism and biomarkers of heterocyclic aromatic amines in molecular epidemiology studies: lessons learned from aromatic amines. Chem Res Toxicol 24, 11691214.Google Scholar
42. Thiebaut, AC, Kipnis, V, Chang, SC et al. (2007) Dietary fat and postmenopausal invasive breast cancer in the National Institutes of Health-AARP Diet and Health Study cohort. J Natl Cancer Inst 99, 451462.Google Scholar
43. Sieri, S, Krogh, V, Ferrari, P et al. (2008) Dietary fat and breast cancer risk in the European Prospective Investigation into Cancer and Nutrition. Am J Clin Nutr 88, 13041312.Google Scholar
44. Smith-Warner, SA, Spiegelman, D, Adami, HO et al. (2001) Types of dietary fat and breast cancer: a pooled analysis of cohort studies. Int J Cancer 92, 767774.Google Scholar
45. Xu, C, Han, FF, Zeng, XT et al. (2015) Fat intake is not linked to prostate cancer: a systematic review and dose-response meta-analysis. PLoS ONE 10, e0131747.Google Scholar
46. zur Hausen, H (2012) Red meat consumption and cancer: reasons to suspect involvement of bovine infectious factors in colorectal cancer. Int J Cancer 130, 24752483.CrossRefGoogle ScholarPubMed
47. Micha, R, Michas, G & Mozaffarian, D (2012) Unprocessed red and processed meats and risk of coronary artery disease and type 2 diabetes–an updated review of the evidence. Curr Atheroscler Rep 14, 515524.Google Scholar
48. Kaluza, J, Wolk, A & Larsson, SC (2012) Red meat consumption and risk of stroke: a meta-analysis of prospective studies. Stroke 43, 25562560.Google Scholar
49. Linseisen, J, Rohrmann, S, Norat, T et al. (2006) Dietary intake of different types and characteristics of processed meat which might be associated with cancer risk–results from the 24-hour diet recalls in the European Prospective Investigation into Cancer and Nutrition (EPIC). Public Health Nutr 9, 449464.Google Scholar
50. He, FJ, Li, J & Macgregor, GA (2013) Effect of longer term modest salt reduction on blood pressure: Cochrane systematic review and meta-analysis of randomised trials. BMJ 346, f1325.Google Scholar
51. Aburto, NJ, Ziolkovska, A, Hooper, L et al. (2013) Effect of lower sodium intake on health: systematic review and meta-analyses. BMJ 346, f1326.Google Scholar
52. O'Donnell, M, Mente, A & Yusuf, S (2015) Sodium intake and cardiovascular health. Circ Res 116, 10461057.Google Scholar
53. Mozaffarian, D, Micha, R & Wallace, S (2010) Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: a systematic review and meta-analysis of randomized controlled trials. PLoS Med 7, e1000252.Google Scholar
54. Chowdhury, R, Warnakula, S, Kunutsor, S et al. (2014) Association of dietary, circulating, and supplement fatty acids with coronary risk: a systematic review and meta-analysis. Ann Intern Med 160, 398406.Google Scholar
55. Hori, A, Mizoue, T, Kasai, H et al. (2010) Body iron store as a predictor of oxidative DNA damage in healthy men and women. Cancer Sci 101, 517522.Google Scholar
56. Nakano, M, Kawanishi, Y, Kamohara, S et al. (2003) Oxidative DNA damage (8-hydroxydeoxyguanosine) and body iron status: a study on 2507 healthy people. Free Radic Biol Med 35, 826832.Google Scholar
57. Iqbal, MP, Mehboobali, N, Tareen, AK et al. (2013) Association of body iron status with the risk of premature acute myocardial infarction in a Pakistani population. PLoS ONE 8, e67981.CrossRefGoogle Scholar
58. Astorg, P, Bertrais, S, Laporte, F et al. (2008) Plasma n-6 and n-3 polyunsaturated fatty acids as biomarkers of their dietary intakes: a cross-sectional study within a cohort of middle-aged French men and women. Eur J Clin Nutr 62, 11551161.Google Scholar
59. Sakai, M, Kakutani, S, Tokuda, H et al. (2014) Arachidonic acid and cerebral ischemia risk: a systematic review of observational studies. Cerebrovasc Dis Extra 4, 198211.Google Scholar
60. Mozaffarian, D, Ascherio, A, Hu, FB et al. (2005) Interplay between different polyunsaturated fatty acids and risk of coronary heart disease in men. Circulation 111, 157164.Google Scholar
61. de Oliveira Otto, MC, Wu, JH, Baylin, A et al. (2013) Circulating and dietary omega-3 and omega-6 polyunsaturated fatty acids and incidence of CVD in the Multi-Ethnic Study of Atherosclerosis. J Am Heart Assoc 2, e000506.Google Scholar
62. Reinders, I, van Ballegooijen, AJ, Visser, M et al. (2013) Associations of serum n-3 and n-6 polyunsaturated fatty acids with echocardiographic measures among older adults: the Hoorn Study. Eur J Clin Nutr 67, 12771283.CrossRefGoogle ScholarPubMed
63. Tang, WH, Wang, Z, Levison, BS et al. (2013) Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med. 368, 15751584.Google Scholar
64. Koeth, RA, Wang, Z, Levison, BS et al. (2013) Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med. 19, 576585.Google Scholar
65. Wang, Z, Klipfell, E, Bennett, BJ et al. (2011) Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature 472, 5763.Google Scholar
66. Hamlin, JC, Pauly, M, Melnyk, S et al. (2013) Dietary intake and plasma levels of choline and betaine in children with autism spectrum disorders. Autism Res Treat 2013, 578429.Google Scholar
67. Rohrmann, S, Linseisen, J, Allenspach, M et al. (2015) Plasma concentrations of trimethylamine-N-oxide are directly associated with dairy consumption and low-grade inflammation in a German adult population. J Nutr (In the Press).Google Scholar
68. Bendinelli, B, Palli, D, Masala, G et al. (2013) Association between dietary meat consumption and incident type 2 diabetes: the EPIC-InterAct study. Diabetologia 56, 4759.Google Scholar
69. Mannisto, S, Kontto, J, Kataja-Tuomola, M et al. (2010) High processed meat consumption is a risk factor of type 2 diabetes in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention study. Br J Nutr 103, 18171822.Google Scholar
70. Suckling, RJ, He, FJ & Macgregor, GA (2010) Altered dietary salt intake for preventing and treating diabetic kidney disease. Cochrane Database Syst Rev CD006763.Google Scholar
71. Tong, M, Neusner, A, Longato, L et al. (2009) Nitrosamine exposure causes insulin resistance diseases: relevance to type 2 diabetes mellitus, non-alcoholic steatohepatitis, and Alzheimer's disease. J Alzheimers Dis 17, 827844.Google Scholar
72. de la Monte, SM, Neusner, A, Chu, J et al. (2009) Epidemilogical trends strongly suggest exposures as etiologic agents in the pathogenesis of sporadic Alzheimer's disease, diabetes mellitus, and non-alcoholic steatohepatitis. J Alzheimers Dis 17, 519529.Google Scholar
73. Orban, E, Schwab, S, Thorand, B et al. (2014) Association of iron indices and type 2 diabetes: a meta-analysis of observational studies. Diab Metab Res Rev 30, 372394.Google Scholar
74. Rajpathak, SN, Crandall, JP, Wylie-Rosett, J et al. (2009) The role of iron in type 2 diabetes in humans. Biochim Biophys Acta 1790, 671681.Google Scholar
75. Peppa, M, Goldberg, T, Cai, W et al. (2002) Glycotoxins: a missing link in the ‘relationship of dietary fat and meat intake in relation to risk of type 2 diabetes in men’. Diab Care 25, 18981899.Google Scholar
76. Malmstedt, J, Karvestedt, L, Swedenborg, J et al. (2015) The receptor for advanced glycation end products and risk of peripheral arterial disease, amputation or death in type 2 diabetes: a population-based cohort study. Cardiovasc Diabetol 14, 93.Google Scholar
77. Thomas, MC, Woodward, M, Neal, B et al. (2015) The relationship between levels of advanced glycation end-products and their soluble receptor and adverse outcomes in adults with type 2 diabetes. Diabetes Care 38, 18911897.Google Scholar
78. Hanssen, NM, Beulens, JW, van Dieren, S et al. (2015) Plasma advanced glycation end products are associated with incident cardiovascular events in individuals with type 2 diabetes: a case-cohort study with a median follow-up of 10 years (EPIC-NL). Diabetes 64, 257265.Google Scholar
79. Montonen, J, Boeing, H, Fritsche, A et al. (2013) Consumption of red meat and whole-grain bread in relation to biomarkers of obesity, inflammation, glucose metabolism and oxidative stress. Eur J Nutr 52, 337345.Google Scholar
80. Ley, SH, Sun, Q, Willett, WC et al. (2014) Associations between red meat intake and biomarkers of inflammation and glucose metabolism in women. Am J Clin Nutr 99, 352360.CrossRefGoogle Scholar
81. Norat, T, Bingham, S, Ferrari, P et al. (2005) Meat, fish, and colorectal cancer risk: the European prospective investigation into cancer and nutrition. J Natl Cancer Inst 97, 906916.Google Scholar
82. Hodek, P, Trefil, P & Stiborova, M (2002) Flavonoids-potent and versatile biologically active compounds interacting with cytochromes P450. Chem Biol Interact 139, 121.Google Scholar
83. Hernandez-Ramirez, RU, Galvan-Portillo, MV, Ward, MH et al. (2009) Dietary intake of polyphenols, nitrate and nitrite and gastric cancer risk in Mexico City. Int J Cancer 125, 14241430.Google Scholar
84. Chan, DS, Lau, R, Aune, D et al. (2011) Red and processed meat and colorectal cancer incidence: meta-analysis of prospective studies. PLoS ONE 6, e20456.Google Scholar
85. Mente, A, de Koning, L, Shannon, HS et al. (2009) A systematic review of the evidence supporting a causal link between dietary factors and coronary heart disease. Arch Intern Med 169, 659669.Google Scholar
Figure 0

Table 1. Association between red and processed meat consumption and all-cause and cause-specific mortality: results of meta-analyses

Figure 1

Table 2. Potential mechanisms connecting meat consumption and risk of chronic diseases