We searched various electronic databases, including PubMed, ProQuest, Scopus, and Web of Science, from inception to March 31, 2022, to identify relevant studies reporting the prevalence of NAFLD and cancers incident among adults (aged ≥18 years). A search strategy was generated systematically for PubMed and adapted for use in the other electronic databases. Keywords and comparable medical subject heading (MESH) terms were used in combination where appropriate, without restriction on language or publication year. Specific search terms were as follows: (“Fatty liver” OR “steatosis”) AND (“cancer” OR “malignancy” OR “tumour” OR “carcinoma”) AND “prevalence” for the PubMed search. Searches were restricted to human studies. Studies in languages other than English were excluded. Additionally, we screened cross-references from relevant original papers and review articles to identify primary studies not covered by the original database searches. We used a reference manager (EndNote; Clarivate Plc) to store all the studies and citations and developed a database for the included and excluded studies. We integrated our reporting with the MOOSE (Meta-Analysis of Observational Studies in Epidemiology) guidelines [23].

We included all observational studies that evaluated the prevalence of NAFLD and examined the association between NAFLD and the risk of developing cancers (Textbox 1). We included studies with an adult population aged 18 years and older of either sex without any restriction in terms of race or ethnicity in the meta-analysis. We excluded conference abstracts, case reports, reviews, commentaries, editorials, practice guidelines, and cross-sectional studies. In the case of multiple studies using the same cohort, the study with the most detailed information on the participants or the largest number of participants was selected. We included studies according to the population, intervention, comparator, outcome, study (PICOS) approach (Table S1 in Multimedia Appendix 3).

Two authors (NHM and FNL) independently screened all the titles and abstracts to search for potential studies identified as a result of the search and coded them as “retrieve” (eligible, potentially eligible, or unclear) or “do not retrieve.” Another 2 authors (ZAM and SKB) independently retrieved the full-text study reports and publications to identify studies for inclusion, as well as identify and record reasons for the studies’ exclusion. In case of discrepancies, an agreement was reached by consensus and discussion with a third reviewer (NML).

Two authors (IAR and MHAM) independently evaluated the methodological quality of each included study and extracted data using an electronic form in the reference manager, which was adapted from the Cochrane Handbook for Systematic Review of Interventions [24]; discrepancies were resolved by discussion with a third author (NAM). A standardized data collection form was used to extract the data. The extraction form contained information about the author, year of publication, setting (country of origin), population, study design, sample size, study period, presence of fatty liver, method of detection of fatty liver, type of cancer, and prevalence of NAFLD.

Methodological quality was determined using a domain-based tool adapted from the Newcastle-Ottawa Scale (NOS) to assess the risk of bias of each study (Multimedia Appendix 4). We classified the risk of bias as either low, moderate, high, or unclear across the following domains: selection of participants (selection bias), sample size justification (selection bias), outcome measurement (detection bias), and confounding adjustment. We assigned a score of 7 and above as good quality, and below 6 as having concerns related to determining the overall quality [25].

We used Stata (version 16; StataCorp) for all statistical analysis. The pooled prevalence rates, as well as their 95% CIs, were calculated using a random effects model [26]. The I² statistic and Cochran Q test were used to assess heterogeneity among the studies [27]. The I² describes the percentage of the variability in effect estimates that is attributable to heterogeneity rather than sampling error. A value greater than 50% may be considered to indicate substantial heterogeneity, whereas a score of more than 75% indicates high heterogeneity [26,28].

From the preliminary search, we identified 4687 studies across all databases, including PubMed (n=3190), ProQuest (n=89), Scopus (n=1290), and Web of Science (n=118). After removing duplicates, we screened a total of 158 titles and abstracts, of which 143 were excluded. We retrieved a total of 15 full texts for inclusion. We excluded 4 studies from the full texts and included 11 studies that met our inclusion criteria for this review. The most common reasons for exclusion were (1) no data on population prevalence of NFALD and cancer was reported and (2) the cancer condition was not included in the study. After removing duplicates, titles, and abstracts, 4 studies were excluded. We included 11 studies, as shown in Table 1 and Figure 1. Figure 2 shows the selection process to complete the PRISMA flow diagram.

A total of 11 studies with 222,523 adult participants (aged ≥18 years) across 6 countries were included. A total of 3 studies were conducted in the United States and 8 studies were from 5 countries in Asia: South Korea (n=3), China (n=2), Turkey (n=1), Japan (n=1), and Israel (n=1). Cancer was classified into 3 types: HCC (5 studies), breast cancer (4 studies), and others (2 studies). The characteristics of the included studies are depicted in Table 1.

We assessed the quality of the included studies [29-39] as NOS quality, with modifications as shown in Table S2 in Multimedia Appendix 3. All 11 studies were observed to have good quality (NOS score 7 and above).

The overall pooled prevalence of NAFLD among the adult population was 26% (95% CI 16%-35%), as shown in Figure 3, with a high level of heterogeneity between studies (I²=99.2%; P=.001). The overall pooled prevalence of HCC was calculated based on these findings, and it was observed that the overall pooled prevalence of HCC was 25% (95% CI 7%-42%) with a high level of heterogeneity between studies (I²=99.6%; P=.001) reported in Asia (4 publications) and the United States (1 publication; Figure 4). The overall pooled prevalence of breast cancer was 30% (95% CI 14%-45%) with a high level of heterogeneity between studies (I²=96.9%; P=.001) reported in Asia (4 publications), as shown in Figure 5. In this review, cancer in first degree relatives and intrahepatic cholangiocarcinoma showed an overall pooled prevalence of 21% (95% CI 12%-31%) with a moderate level of heterogeneity between studies (I²=71%; P=.001) reported in Asia and the United States, each with 1 publication (Figure 6).

This review elucidates an assessment of NAFLD as a worldwide epidemic that may contribute to chronic liver disease. In this systematic review and meta-analysis, we included 11 studies to comprehensively estimate the global prevalence and the association with cancers among the adult population with NAFLD. There is strong evidence that suggests the association of NAFLD with various diseases, such as cardiovascular disease, diabetes, and cancers such as bile duct, breast, and liver cancer [11,40]. In recent years, more research and discussion has been focused on the possible association between NAFLD and the risk of cancers. Therefore, this review will be useful in expounding the latest research progress on this issue [41].

In this review, the overall pooled prevalence of NAFLD and cancers was 26%. The pooled prevalence of NAFLD and breast cancer was higher than that of HCC and other cancers (30% vs 25% and 21%, respectively). This review also discovered that the highest prevalence of NAFLD is reported in breast cancer, which is consistent with the findings of studies by Kim et al [42] (2017) and Mantovani et al [43] (2022). According to the former report, NAFLD showed a strong association with the development of HCC and colorectal cancer in men and breast cancer in women [42]. The latter study, a meta-analysis, examined the risk of several prespecified cancers and found an increased risk of developing breast cancer [43].

A review by Thomas et al [44] (2022) examined NAFLD and the incidence of hepatic and extrahepatic cancers. The review found that the pooled incidence rate of HCC was 1.25 per 1000 person-years. However, our review examined the prevalence of HCC among the adult population with NFLD. Another review [45] found that the presence of NAFLD was independently associated with an 88% increased risk of HCC, as compared to the absence of NAFLD. This review is different from ours, which examined the prevalence of HCC with NAFLD.

Although the exact or detailed mechanism of this interaction remains unclear, cancer development in patients with NAFLD may be associated with a bidirectional interaction between NAFLD and metabolic syndrome [39]. According to Tiniakos et al [46] (2018), the increased susceptibility of the liver with steatosis to carcinogenic insults could be linked to metabolic derangements, such as metabolic syndrome, hyperinsulinemia, and the presence of insulin-like growth factor receptors in HCC, as well as systemic effects of deranged cytokines and adipokines, immune dysregulation, and changes in gut microbiota. Besides that, the genetic component has been named as another factor that contributes to an increased risk of HCC in individuals with NAFLD [46].

A few mechanisms for extrahepatic carcinogenesis of the fatty liver, such as breast cancer, have been proposed. First, high levels of inflammatory cytokines are closely associated with NAFLD, as they promote insulin resistance and elevated circulating triglycerides, influence growth, and increase apoptosis and tumor cell proliferation in many cancers [47]. Second, hyperinsulinemia and high levels of leptin have carcinogenic effects [48]. By binding to the circulating sex hormone–binding globulin, increasing insulin levels cause the elevated secretion of estrogen, and downstream signaling favors breast carcinogenesis [49]. Insulin may cross-bind to insulin-like growth factor-I (IGF-1) receptors on breast cells, and downstream signaling pathways provide proliferation stimuli to breast cancer cells [19]. Third, decreased levels of adiponectin lead to marked insulin resistance and subsequent increased levels of IGF-1. Insulin binds to IGF-1 receptors and plays an important role in cell proliferation, apoptosis, and increased production of vascular endothelial growth factors [37].

This study has several strengths. To our knowledge, this is a comprehensive and up-to-date systematic review and meta-analysis that reports the pooled prevalence of NAFLD and cancers. We examined a number of studies using stringent inclusion criteria. Following the broad search strategy, we were able to stratify hepatic and extrahepatic cancers. There are potential limitations in our review. The included studies largely had an observational design, and there are risks for a number of biases in term of design, selection of respondents, and sample size. The included studies were located only in Asia (n=8) or the United States (n=3), limiting the accuracy of the estimations. There is a paucity of data from other countries, and most studies are conducted among the urban population. Therefore, our findings might be biased and have limited generalizability beyond the countries included in this review. In short, there is a need to investigate NAFLD and cancer across different regions in the future.

This systematic review and meta-analysis shows that the pooled prevalence of NAFLD is closely associated with hepatic and extrahepatic cancers. This review provides evidence that a substantially high proportion of patients with NAFLD are associated with extrahepatic and hepatocellular carcinoma. However, the evidence remains inconclusive, and further studies are needed to confirm the association between NAFLD and cancers, as well as to improve surveillance strategies for patients with NAFLD who are at high risk of cancers. Since the global prevalence of NAFLD is increasing, policymakers must work toward reversing the current trends by increasing the awareness of NAFLD and promoting healthy lifestyles and environments.