Acrylamide is a chemical compound mainly used in industry to produce polyacrylamides that are employed for different purposes such as flocculants, dyes, paper, and textiles. Beside the potential exposure of acrylamide in the occupational setting, the main sources of exposure to this substance in the general population is through the intake of certain foods, high in carbohydrates cooked at temperatures >120°C, especially in low moisture conditions (1). Acrylamide may also be formed in foods from acrolein, a compound produced during the degradation of amino acids, carbohydrates, lipids, and organic acids (2). The content of acrylamide in foods varies largely, depending on the food matrix and the food processing method. Thus, dietary exposure to acrylamide differs across countries and global geographical areas, due to distinct traditional diets and its culinary habits (3). Tobacco smoke is also a source of acrylamide, making smokers potentially exposed to higher levels of acrylamide compared to non-smokers (4, 5). Smokers have been shown to have, on average, three to four times higher levels of acrylamide biomarkers (i.e., hemoglobin adducts) compared to non-smokers (6). Additional sources of exposure to acrylamide take place in occupational settings, where an association to neurological alterations has been shown (7).

Acrylamide is classified as probably carcinogenic to humans (class 2A) by the International Agency for Research on Cancer (IARC) (8). However, the underlying mechanisms of the carcinogenicity of acrylamide are far from being understood. Still, experimental evidence has shown that the main metabolite of acrylamide, glycidamide, is genotoxic—highly reactive toward DNA and proteins (9, 10). A study found a glycidamide-related mutational signature in one-third of approximately 1,600 human tumor genomes corresponding to 19 types of cancer (11). Non-genotoxic modes of action of acrylamide involving, for instance, hormonal pathways, are also proposed as underlying mechanisms of carcinogenicity, driving cell transformation or proliferation and apoptosis, independently of acrylamide-induced genotoxicity (9).

Although in vitro and animal studies have consistently shown that acrylamide is genotoxic, mutagenic, and carcinogenic (9, 12), epidemiological studies investigating the relationship between high dietary acrylamide exposure and the risk of several site-specific cancer (e.g., prostate, gastrointestinal and lung) in the general population and in occupational settings have reported inconsistent results, with some studies pointing to an increased risk (13–19) and others to null association (3, 17, 18, 20–41). All of these studies used a dietary assessment method to estimate dietary acrylamide exposure except for one study performed in a Swedish population where acrylamide exposure was assessed through both dietary intake and hemoglobin adducts (20). So far, a few systematic reviews and meta-analyses based on epidemiological studies were performed to synthesize the body of evidence in this field (42–48). In the most recent and comprehensive meta-analysis published in 2015, no association between high dietary acrylamide exposure and the risk of several cancers was noted, except for a modest increased risk of renal cancer (46), which was not observed in another meta-analysis investigating dietary acrylamide exposure and renal cell carcinoma (44). With the exception of a recent study investigating the relationship between dietary acrylamide exposure and female reproductive cancers, observing a relatively linear increased risk for ovarian and endometrial cancer (42), no dose-response meta-analysis exist for other cancer forms.

Hence, we aimed to synthesize the results on the association between dietary acrylamide exposure and risk of site-specific cancer. For this purpose, we performed a systematic literature review and dose-response meta-analysis of epidemiological studies evaluating the association between acrylamide and the risk of several site-specific cancer.

Out of 1,994 papers retrieved through the literature and manual search, we identified 31 papers that met the inclusion criteria, corresponding to a total of 16 studies. Detailed information on the selection of the studies included is presented in Figure 1. Descriptive characteristics of the studies included are presented in Table 1. Most studies were performed in Europe (n = 21), primarily in the Netherlands (n = 8) and Sweden (n = 7), followed by Japan (n = 6) and the United States (n = 3). A total of 1,151,189 participants were included in these case-control, case-cohort, and cohort studies. Participants’ age ranged from 50 to 70 years (mean: 60 years). During a median follow-up period of 14.9 years (range: 7.3–33.9 years), 48,175 participants developed cancer. The following site-specific cancer were investigated in the papers in relation to dietary acrylamide exposure: oral cavity, esophageal, stomach, colorectal (including colon and rectal), pancreatic, laryngeal, lung, lymphoma, multiple myeloma, renal, bladder/urothelial, prostate, melanoma, and brain. Dietary acrylamide was assessed through food frequency questionnaires in all the analyzed studies. The mean and median estimated dose of dietary acrylamide across studies was 23 μg/day.

Summary estimates for the association between highest vs. lowest dietary acrylamide exposure and the risk of various types of cancer in smokers and never smokers combined, and by smoking status, are shown in Table 2 and in forest plots (Supplementary Figures 1–15). Regardless smoking status, summary estimates showed no association between higher dietary acrylamide and any of the site-specific cancer investigated.

Results from the dose-response analysis showed a null association with no thresholds for all site-specific cancer considered in relation to increasing levels of dietary acrylamide (Figures 2, 3).

There was also no association between dietary acrylamide and the risk of site-specific cancer in the subgroups of smokers and never smokers, except for lung cancer, where we observed an increased risk in smokers (RR 1.16; 95% CI 1.03–1.31). Compared to smokers, the summary estimates for never smokers were observed to be higher, albeit very imprecise, for stomach, pancreatic, liver, renal, prostate, and multiple myeloma (Supplementary Figures 3, 7, 11, 12, 14).

Additional analysis by geographical regions (West and East) could be performed only excluding Asian population (i.e., Japan studies) since most of the studies were performed in the West area (Europe and United States). We observed a slightly, though imprecise, increased risk of lymphoma in relation to high dietary acrylamide when we consider only study from the West (n = 1) (RR 1.12; 95% CI 0.99–1.22) (Supplementary Figure 13).

There was no evidence of between-study heterogeneity except for lung and multiple myeloma cancers (Supplementary Figures 9, 14, respectively) (I² > 50%). The graphical overlay of single study effects in the dose-response meta-analysis did not show substantial variation (data not shown). No evidence of publication bias, visualized by funnel plots, was observed (Supplementary Figures 2, 4, 7, 11, 12). The results of the risk of bias assessment, using the ROBINS-I are shown in Supplementary Figures 16, 17. Within the six dimensions analyzed, all studies had a low risk of bias regarding the measurement of outcomes and the reported result, and 75% of the studies showed a low risk of bias on the dimensions selection of participants and classification of interventions. All studies had a moderate risk of bias for confounding. There was missing information (classified as “no information”) regarding the dimensions’ deviations from intended interventions and missing data for more than 75% of the analyzed studies.

In this up-to-date systematic review and meta-analysis, based on 31 papers, we observed no association between high dietary acrylamide exposure and increased risk of any of the investigated cancers, including those of oral cavity, esophageal, gastric, colon-rectal, pancreatic, prostate, bladder, lung, renal, lymphoma, myeloma, thyroid, brain, larynx, and melanoma. As a novel finding, we found no thresholds between different levels of dietary acrylamide and the risk of any of the site-specific cancers considered. Considering studies performed in Western geographical areas alone, a slightly increased risk of lymphoma was observed in relation to high dietary acrylamide exposure. In general, findings did not differ by smoking status, except for increased risk of lung cancer in smokers.

Overall, our results of no association between high dietary acrylamide exposure and different site-specific cancers in the general population confirm the findings of previous meta-analyses that did not perform dose-response meta-analysis (22, 44, 46, 48). On the other hand, Pelucchi et al. (46) found a slight increased risk of renal cell carcinoma by 20% in relation to high dietary acrylamide (RR 1.20; 95% CI 1.00–1.45). This difference may be explained by the fact that we included three recent cohort studies with American (18, 30) and Japanese (59) participants. Results from these additional studies were to some extent heterogenous, with either a tendency of increased (18, 30) or decreased risk (30, 59) of renal cell carcinoma. The studies with American participants (18, 30) also contributed largely to the analysis (around 10% weight), possibly affecting our overall results. Moreover, we did not include the Finnish Alpha-Tocopherol, Beta-Carotene Cancer (ATBC) prevention study (60), a cohort study of 27,111 smoking men, in our main analysis which was based on both smokers and never smokers.

To the best of our knowledge, our meta-analysis is the first one investigating the shape of the dose-response relationships between dietary acrylamide exposure and the risk of site-specific cancer. Our results suggest that these associations, if present, may generally be without thresholds. These results are in line with those reported in a previous meta-analysis investigating the association between dietary acrylamide exposure and female reproductive cancers, which were not considered for inclusion in our meta-analysis (42).

Our aforementioned results may be explained by the relatively low dietary acrylamide exposure in the general population (intake ranged from 6.8 to 44.1 μg/day in the included studies—corresponding up to about 0.7 μg/kg body weight/day) when compared to the levels of acrylamide observed to be toxic in animal studies (50 mg/kg body weight/day) (9). There is also evidence showing that acrylamide does not generate any toxicologically detrimental effects when male rats were administrated three low oral doses of acrylamide (20, 40, and 90 μg/kg body weight/day, respectively) (61). Also, it is relevant to take into account that the EFSA CONTAM Panel selected BMDL₁₀ value of 0.17 mg/kg body weight/day for neoplastic effects in mice (12), i.e., much higher level compared with those generally experienced by humans. Most of the included epidemiological studies investigating the relationship between dietary acrylamide and cancer risk were performed in North European regions, where dietary exposure to acrylamide is higher than in the Eastern geographical regions (i.e., Japanese populations), which reported the lowest mean dietary exposure below 10 μg/day. Additionally, the no association found in this study may be explained by the lack of assessment of co-exposure to other compounds in food matrixes, including other foodborne carcinogens (62) or substances that can confer protection to carcinogenic compounds (63, 64).

The findings related to the slightly increased risk of lymphomas in relation to high dietary acrylamide exposure restricted to the Western geographical region needs to be interpreted with caution as they are based on one single study (16) and combining the subtypes of lymphoma, characterized by very different clinical features, as well as genetic profile (65). These findings disagree to some extent with those reported by Pelucchi et al. (46). However, the latter did not investigate lymphomas and multiple myeloma separately and considered the ATBC study that we excluded from the main analysis because based on smokers alone. The biological mechanisms underlying the potentially increased risk of lymphoma in relation to high dietary acrylamide exposure in Western regions are unclear, but explanations may be speculated. Firstly, the hydrophilic nature of acrylamide leads to systemic exposure and so all human tissues, including the hemopoietic and lymphoid tissues, are potential targets for acrylamide-induced carcinogenesis. Also, acrylamide has been shown to induce carcinogenicity through dysregulation of the endocrine system, including hormones (66) and this has been suggested to play a role in the pathophysiology of lymphomas (16). The restriction of these findings to the Western geographical areas may be attributed to possible interactions between genetic and environmental exposures determinants, such as different sources of dietary acrylamide (67, 68).

In general, the similar results found in smokers and never smokers in our meta-analysis are in line with a previous meta-analysis (46). The results of increased risk of lung cancer in smokers and melanoma in non-smokers need be carefully interpreted since they are based on few and heterogenous studies. Also, for some of the cancers considered, the summary estimates suggested that never smokers could experience a slightly positive association between dietary acrylamide and cancer risk, which was not detectable in smokers. These results are in line with previous studies investigating this relationship by smoking status (16, 60). A possible explanation is that for the same level of exposure, less dietary acrylamide is converted to the genotoxic and carcinogenic metabolite of acrylamide, glycidamide, because smoking affects the metabolism of acrylamide by e.g., saturating enzymes involved in its conversion to glycidamide (55).

Based on the results of this dose-response meta-analysis of epidemiological studies, higher dietary acrylamide exposure (vs. lower) was not associated with an increased risk of several site-specific cancers. If associations between dietary acrylamide and the risk of several site-specific cancers were present, the shape of the dose-response relationships would be with no thresholds. Smoking status might modify the relation between dietary acrylamide and some site-specific cancers, but findings need to be furtherly investigated.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

TF, TH, AV, and FL conceptualized the study and drafted the manuscript. TF performed the statistical analyses. All authors contributed substantively to this work, involved in the interpretation of results, editing of the manuscript, and approved manuscript submission.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.