Next Article in Journal
Screening Colonoscopy Uptake Among Adult Stroke Survivors: Findings from the 2022 BRFSS Data
Previous Article in Journal
Drug-Induced Liver Injury: Role of Circulating Liver-Specific microRNAs and Keratin-18
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Gastroenterology Insights
Volume 16
Issue 1
10.3390/gastroent16010001
gastroent-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Smoking and Risk of Fatty Liver Disease: A Meta-Analysis of Cohort Studies

by
Moonhyung Lee
1,
Seung-Kwon Myung
2,3,4,*,
Sang Hee Lee
5 and
Yoosoo Chang
6
1
Department of Gastroenterology and Hepatology, Kyung Hee University Hospital at Gangdong, Kyung Hee University College of Medicine, Seoul 05278, Republic of Korea
2
Department of Family Medicine, Hospital, National Cancer Center, Goyang 10408, Republic of Korea
3
Department of Cancer AI & Digital Health, National Cancer Center, Graduate School of Cancer Science and Policy, Goyang 10408, Republic of Korea
4
Cancer Epidemiology Branch, Division of Cancer Data Science, Research Institute, National Cancer Center, Goyang 10408, Republic of Korea
5
Department of Nephrology, Veterans Health Service Medical Center, Seoul 05368, Republic of Korea
6
Center for Cohort Studies, Total Healthcare Center, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul 03181, Republic of Korea
*
Author to whom correspondence should be addressed.
Gastroenterol. Insights 2025, 16(1), 1; https://doi.org/10.3390/gastroent16010001
Submission received: 30 October 2024 / Revised: 20 December 2024 / Accepted: 31 December 2024 / Published: 2 January 2025
(This article belongs to the Section Liver)
Browse Figures
Review Reports Versions Notes

Abstract

:
Background/Objectives: It remains inconclusive whether or not smoking is associated with an increased risk of fatty liver disease (FLD). We investigated the association between smoking and the risk of FLD by using a meta-analysis of cohort studies. Methods: PubMed and EMBASE were searched using keywords from their inception to September 2023 to identify relevant studies. Results: Out of 806 articles searched from the databases, a total of 20 cohort studies were included in the final analysis. In the meta-analysis, smoking was significantly associated with an increased risk of FLD (odds ratio/relative risk/hazard ratio, 1.14; 95% confidence interval, 1.05–1.24; n = 20). Subgroup analyses showed a significant positive association between them in prospective cohort studies (odds ratio/relative risk/hazard ratio, 1.15; 95% confidence interval, 1.05–1.18; n = 5) but not in retrospective cohort studies and cross-sectional studies based on cohort studies. In the subgroup meta-analysis by gender in Asians, smoking significantly increased the risk of FLD in men, while there was no significant association between FLD and smoking in women. Conclusions: This meta-analysis showed that smoking increases the risk of FLD. In addition to well-known risk factors of FLD, clinicians should recommend smoking cessation for the management of FLD.

1. Introduction

Fatty liver disease (FLD), which is categorized into two major types, namely nonalcoholic FLD (NAFLD) or alcoholic FLD (AFLD), indicates a morphological spectrum consisting of hepatic steatosis and steatohepatitis [1]. It is a persistent liver condition marked by macrovesicular steatosis in liver cells, which has the potential to advance to hepatic cirrhosis, liver failure, and possibly hepatocellular carcinoma [1]. Ethanol consumption serves as a pivotal determinant in distinguishing between NAFLD and AFLD in the guidelines from the European Association for the Study of the Liver [2]. The diagnosis of NAFLD requires an ethanol intake of 20 g/d or less in females and 30 g/d or less in males following the thorough exclusion of alternative causes, such as hepatitis virus infection and the use of steatogenic drugs [3]. Beyond alcohol-related impacts, the pathogenesis of FLD is intricately linked to various contributors, including insulin resistance (IR), oxidative stress, mitochondrial dysfunction, immune system deregulation, and the release of adipokines [3,4]. These factors collectively underscore the complexity of FLD, emphasizing the importance of a comprehensive understanding to address its progression and associated risks, such as hepatic cirrhosis, liver failure, and hepatocellular carcinoma [5].
The overall prevalence of NAFLD worldwide also has increased from 25.5% in or before 2005 to 37.8% in 2016 or later based on a report from a recent meta-analysis published in 2022 [6]. Also, FLD has become a predominant chronic liver disorder in developed Western countries [7]. Its risk factors include a higher body mass index (BMI), consumption of saturated fat and fructose, type 2 diabetes, and known single-nucleotide polymorphisms [8].
However, it remains unclear whether smoking is associated with an increased risk of FLD. An animal study has reported that smoking increases lipid accumulation in hepatocytes by modulating an activity of critical molecules related with lipid synthesis [9]. Also, another animal study has shown that the histological severity of NAFLD in obese rats was exacerbated by tobacco exposure [10].
In the meantime, observational epidemiological studies [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30] have reported inconsistent findings. Several cohort studies [14,17,18,20,23,24,27,28,29] reported that smoking was associated with an increased risk of FLD, whereas other cohort studies [11,12,13,15,16,19,22,26] indicated no significant association between them.
In 2018, the only conducted meta-analysis of observational studies reported that smoking was significantly associated with an increased risk of NAFLD [31]. However, it included a small number of cohort studies to confirm the association, and subsequent cohort studies have been published since then. Furthermore, to our knowledge, no meta-analysis of cohort studies has been published regarding the association between smoking and the risk of FLD encompassing NAFLD and AFLD.
This study aimed to explore the associations between smoking and the risk of FLD by using a comprehensive meta-analysis of cohort studies and subgroup meta-analyses based on important factors.

2. Materials and Methods

2.1. Search Strategy

PubMed and EMBASE were searched in September 2023 with terms obtained from the National Library of Medicine (NLM) Medical Subject Headings (MeSH) and commonly used keywords. We used a PICO framework to combine search terms: P for population is any type of population; I for intervention is ‘smoking’; C for comparison is ‘non-smoker’; and O for outcome is ‘fatty liver disease’. Also, the study type was confined to cohort studies. Thus, the final search terms were ‘smoking’, ‘fatty liver disease’, and ‘cohort study’.

2.2. Study Selection and Data Extraction

We included a cohort study that explored the associations between smoking and FLD (NAFLD or AFLD) and presented risk estimates such as odds ratio (OR), relative risk (RR), or hazard ratio (HR) with their corresponding 95% confidence intervals (CIs). Two independent authors (MH. Lee and SH. Lee) conducted a selection of relevant studies by reviewing titles and abstracts. Discrepancies between them were resolved through discussion. The extracted information included the last name of the first author, publication year, study region, study design (prospective or retrospective cohort study), gender, study participants, comparison of exposure, risk estimates (OR, RR, and HR with corresponding 95% CIs), type of outcomes, and adjusted variables.

2.3. Assessment of Methodological Quality

We used the Newcastle–Ottawa Scale (NOS) in order to assess the methodological quality of the cohort studies included in the current meta-analysis [32]. The NOS comprises eight items and provides a scoring system ranging between 0 and 9. We classified individual cohort studies as having a high or low quality based on the mean score.

2.4. Main and Subgroup Analyses

We investigated the association between smoking and the risks of FLD for the main analysis. We also conducted subgroup meta-analyses by type of study (prospective or retrospective cohort study), region (Europe, Asia, or US), type of FLD, gender (male or female), follow-up period (<5 years or >5 years), and study quality (high or low quality).

2.5. Statistical Analysis

A combined OR, RR, or HR with its corresponding 95% CIs was calculated utilizing the adjusted OR, RR, or HR and their respective 95% CIs from each study that reported the association between smoking and the risk of FLD. The DerSimonian and Laird method [33] was employed, opting for a random-effects model due to the diverse populations across studies. Heterogeneity was evaluated using Higgins I2 and computed as follows:
I2 = 100% × (Q − df)/Q,
where Q is Cochran’s statistic for heterogeneity and df is degrees of freedom [32]. The I2 values range between 0% (no heterogeneity) and 100% (maximal heterogeneity) [32]. Publication bias was assessed using both Begg’s funnel plot and an Egger’s test. We used the STATA SE version 15.1 software package (StataCorp, College Station, TX, USA) for statistical analyses.

3. Results

3.1. Study Selection

Figure 1 shows the diagram of identifying relevant studies. A total of 806 studies were identified by the initial search of the PubMed and EMBASE databases by using keywords. After excluding duplicates, 622 articles were screened based on the review of each title and abstract. After excluding 546 articles according to the predetermined selection criteria, 76 full-text articles were reviewed and assessed for eligibility. Among them, 56 articles were excluded for the following reasons: not relevant (n = 46); not cohort studies (n = 4); identical population (n = 4); and insufficient data (n = 2). The final analysis included 20 cohort studies (Figure 1) [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30].

3.2. General Characteristics of Included Studies

Table 1 shows the general characteristics of the cohort studies included in the final analysis. Eight studies were conducted in Europe, 10 were conducted in Asia, and the remaining two studies were conducted in the US. The types of study designs include prospective cohort studies (n = 5), retrospective cohort studies (n = 6), and cross-sectional surveys based on cohort studies (n = 9). Types of outcomes are FLD (n = 6) and NAFLD (n = 14).

3.3. Methodological Quality of Studies

The average score of the 11 prospective and retrospective cohort studies assessed using the NOS was 7.5 (Table 2). Each study’s methodological quality was categorized as either high (a score of ≥8) or low (a score of <8). Out of the 11 studies, 6 were classified as high-quality studies, while the remaining 5 were categorized as low-quality studies.

3.4. Association Between Smoking and Risk of FLD

In the meta-analysis of all the included studies, smoking was associated with an increased risk of FLD (OR/RR/HR = 1.14; 95% CI, 1.05–1.24; n = 20) (Figure 2).

3.5. Subgroup Meta-Analyses

Table 3 shows the associations between smoking and the risk of FLD in the subgroup meta-analysis by various factors. In the subgroup meta-analysis by study design, smoking increased the risk of FLD in prospective cohort studies (HR = 1.15; 95% CI 1.05–1.18; n = 5) but not in retrospective cohort studies (OR/RR = 1.23; 95% CI 0.94–1.62) and cross-sectional surveys (OR/RR = 1.12; 95% CI 0.92–1.46; n = 9). In the subgroup meta-analysis by study region, smoking was significantly associated with an increased risk of FLD in Europe (OR/RR/HR = 1.32; 95% CI 1.16–1.50; n = 8) but not in Asia (OR/RR/HR = 1.03; 95% CI 0.91–1.18; n = 10) or the US (OR/RR/HR = 0.75; 95% CI 0.28–2.06; n = 2). In the subgroup meta-analysis by gender, for which the only available data were gathered from Asian studies, smoking significantly increased the risk of FLD in men (OR/RR/HR = 1.15; 95% CI 1.06–1.25; n = 4) and there was no significant association between FLD and smoking in women (OR/RR/HR = 1.12; 95% CI 0.94–1.34; n = 4).
Regardless of type of FLD, smoking consistently increased the risk of FLD. On the contrary, subgroup meta-analyses examining the follow-up period and study quality showed no significant association between smoking and the risk of FLD.

3.6. Publication Bias

Both the Begg’s funnel plots (Figure 3) and Egger’s test (p = 0.29) did not show publication bias.

4. Discussion

In this meta-analysis of cohort studies, we found that smoking was significantly associated with an increased risk of FLD. In the subgroup meta-analysis by study design, smoking increased the risk of FLD in prospective cohort studies but not in retrospective cohort studies and cross-sectional surveys. Also, smoking was significantly associated with an increased risk of FLD in Europe but not in Asia or the US. Interestingly, in the subgroup meta-analysis by gender in Asians, smoking significantly increased the risk of FLD in men, while there was no significant association between FLD and smoking in women.
There are several possible biological mechanisms that could explain the increased risk of FLD by smoking. First, Yuan et al.’s study demonstrated that smoking stimulated lipid accumulation in hepatocytes in mice and cultured hepatocytes [9]. When mice and cultured hepatocytes were exposed to sidestream smoke, lipid accumulation was increased by modulating the activity of AMP-activated protein kinase and sterol response element-binding protein-1, which are critical molecules in lipid synthesis [9]. Second, smoking has the potential to induce insulin resistance, which leads to the development of NAFLD [34,35]. A study revealed that cigarette consumption was correlated with the degree of insulin resistance in smokers [36]. Also, a study of biopsy-proven NAFLD patients and health control subjects showed that the homeostasis model assessment of insulin resistance (HOMA-IR) levels was significantly higher in the NAFLD patients [35]. Insulin resistance induced by smoking could enhance hepatic fat accumulation through increasing free fatty acid delivery to the liver and through hyperinsulinemia to contribute to triacylglycerol accumulation in the liver [37]. Third, nicotine in tobacco smoke could increase the release of norepinephrine and epinephrine, which could affect thermogenesis in adipose tissue, leading to the increased lipolysis and the subsequent recycling of fatty acids into triglycerides [38]. Consequently, it may contribute to the development of NAFTD. Lastly, adiponectin could inhibit liver fat deposition, and gluthathione peroxidase (GPx) could reduce lipid and hydrogen peroxide [39]. Thus, it has been suggested that cigarette smoking, in combination with single-nucleotide polymorphisms in the adiponectine gene and GPx1 gene mutation, could contribute to the development of NAFLD [39].
Previously, a meta-analysis of 12 observational studies by Akhavan et al. already reported that smoking moderately increased the risk of NAFLD [31]. However, it included only two cohort studies, and the remaining included studies were seven cross-sectional and three case–control studies. It is possible to combine different study designs, such as cross-sectional, case–control, and cohort studies, when conducting a meta-analysis. However, conducting subgroup meta-analyses by study design is crucial and useful because there could be discrepancies in findings between different study designs. Also, based on the ‘levels of evidence pyramid’, cohort studies generally provide us a higher level of evidence than cross-sectional and case–control studies [40]. Akhavan et al. performed subgroup meta-analyses by study design and reported a significantly increased risk of NAFLD by smoking in all the subgroup meta-analyses by each study design, including cross-sectional, case–control, and cohort studies [31]. However, they included just two cohort studies, which are not enough to confirm the association between smoking and the risk of NAFLD. In the current meta-analysis, we included 11 cohort studies, involving five prospective and six retrospective cohort studies, and nine cross-sectional surveys based on cohort studies. Thus, our findings would provide more clear and convincing evidence on this topic.

Strengths and Limitations

A notable strength lies in the subgroup meta-analyses by various factors. We confirmed a significantly increased risk of FLD in prospective cohort studies, while retrospective and cross-sectional studies showed no significant association between smoking and the risk of FLD.
Interestingly, a significantly increased risk of FLD by smoking was observed in the studies conducted in Europe but not Asia and the US. We included just two studies conducted in the US, and so there is not enough information available to draw a definite conclusion. Thus, more studies are required to confirm the association for people in the US. On the contrary, the number of studies conducted in Asia is sufficient enough to draw a conclusion. We do not have exact reasons why there were discrepancies in findings on this topic between Europeans and Asians. However, a potential explanation might be considered. A misclassification of Asian women’s smoking status might lead to a non-significant association between smoking and the risk of FLD in the current analysis. Unlike adult men, smoking rates among adult women have been known to be very low (<10%) in South Korea, China, and Hong Kong [41,42]. One of the main reasons for the low smoking prevalence in Asian women is under-reporting. The accuracy of smoking rates self-reported by women in Asian countries is doubted because social repression and disapproval of women’s smoking might make women reluctant to report their smoking status [43,44,45]. A study using the Korean National Health and Nutrition Examination Survey reported that, among the cotinine-verified smokers, about 60% of women classified themselves as non-smokers in a self-report survey [42]. When we performed a subgroup analyses by gender, smoking significantly increased the risk of FLD in Asian men, although there was no significant association between smoking and FLD in Asian women. Thus, the discrepancy in findings between Europeans and Asians might be mainly attributable to a misclassification of Asian women’s smoking status but not race or ethnicity. Additionally, differences in diet, genetics, and metabolic profiles, as well as the reliance on self-reported data in many Asian studies, may contribute to attenuating the observed association.
Despite the strengths, there are several limitations in this study. First, we included only five prospective cohort studies. Out of 20 cohort studies included in the current analysis, most studies were retrospective cohort studies and cross-sectional surveys based on cohort studies. Our findings should be confirmed by further prospective cohort studies. Second, most studies measured smoking status based on a self-report. As discussed earlier, the self-reported measure of smoking status may lead to under-reporting and a misclassification of smoking status, which could result in biased conclusions. Thus, further studies with a biochemical validation of smoking status, such as urinary cotinine levels, are warranted to confirm our findings. Third, while the focus of our study was primarily on NAFLD and smoking’s association with fatty liver diseases, we acknowledge the emerging relevance of Metabolic Alcohol-Related Liver Disease (MetALD). However, our inclusion criteria did not separate data specific to MetALD. Future research should address this gap by distinguishing MetALD cases from other subtypes of liver diseases in similar analyses. Fourth, the included studies in our meta-analysis did not provide sufficient data on the quantity of cigarettes smoked and their direct association with NAFLD risk. Most studies categorized participants into smokers and non-smokers without a detailed stratification by smoking intensity. This limitation prevents us from exploring dose–response relationships, which future studies should investigate in order to better understand the impact of smoking intensity on FLD development. Last, we were unable to investigate the association between the number of cigarettes smoked and the risk of FLD due to a lack of data reported in each study.
In summary, we found that smoking increased the risk of FLD in the meta-analysis of cohort studies. In addition to well-known risk factors of FLD, clinicians should recommend smoking cessation for the management of FLD. Furthermore, our findings highlight the need for public health initiatives targeting smoking cessation as part of comprehensive liver disease prevention strategies. Future research should focus on dose–response relationships, exploring the molecular pathways linking smoking to FLD progression and investigating potential interventions tailored to populations who are at higher risk. These steps are essential to advancing both clinical and epidemiological understandings of FLD management.

Author Contributions

M.L.: Conceptualization, data curation, formal analysis, investigation, methodology, visualization, writing—original draft, and writing—review and editing. S.-K.M.: Formal analysis, investigation, methodology, project administration, validation, writing—original draft, and writing—review and editing. S.H.L.: Data curation, methodology, and writing—review and editing. Y.C.: Methodology, writing—original draft, and writing—review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no financial support to produce this manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

No potential conflicts of interest are disclosed.

References

  1. Adams, L.A.; Lymp, J.F.; St Sauver, J.; Sanderson, S.O.; Lindor, K.D.; Feldstein, A.; Angulo, P. The natural history of nonalcoholic fatty liver disease: A population-based cohort study. Gastroenterology 2005, 129, 113–121. [Google Scholar] [CrossRef] [PubMed]
  2. Nascimbeni, F.; Pais, R.; Bellentani, S.; Day, C.P.; Ratziu, V.; Loria, P.; Lonardo, A. From NAFLD in clinical practice to answers from guidelines. J. Hepatol. 2013, 59, 859–871. [Google Scholar] [CrossRef]
  3. Crabb, D.W.; Galli, A.; Fischer, M.; You, M. Molecular mechanisms of alcoholic fatty liver: Role of peroxisome proliferator-activated receptor alpha. Alcohol 2004, 34, 35–38. [Google Scholar] [CrossRef]
  4. Machado, M.; Cortez-Pinto, H. Non-alcoholic steatohepatitis and metabolic syndrome. Curr. Opin. Clin. Nutr. Metab. Care 2006, 9, 637–642. [Google Scholar] [CrossRef] [PubMed]
  5. Machado, M.V.; Cortez-Pinto, H. Management of fatty liver disease with the metabolic syndrome. Expert Rev. Gastroenterol. Hepatol. 2014, 8, 487–500. [Google Scholar] [CrossRef]
  6. Riazi, K.; Azhari, H.; Charette, J.H.; Underwood, F.E.; King, J.A.; Afshar, E.E.; Swain, M.G.; Congly, S.E.; Kaplan, G.G.; Shaheen, A.A. The prevalence and incidence of NAFLD worldwide: A systematic review and meta-analysis. Lancet Gastroenterol. Hepatol. 2022, 7, 851–861. [Google Scholar] [CrossRef] [PubMed]
  7. Browning, J.D.; Szczepaniak, L.S.; Dobbins, R.; Nuremberg, P.; Horton, J.D.; Cohen, J.C.; Grundy, S.M.; Hobbs, H.H. Prevalence of hepatic steatosis in an urban population in the United States: Impact of ethnicity. Hepatology 2004, 40, 1387–1395. [Google Scholar] [CrossRef]
  8. Ko, E.; Yoon, E.L.; Jun, D.W. Risk factors in nonalcoholic fatty liver disease. Clin. Mol. Hepatol. 2023, 29, S79–S85. [Google Scholar] [CrossRef]
  9. Yuan, H.; Shyy, J.Y.; Martins-Green, M. Second-hand smoke stimulates lipid accumulation in the liver by modulating AMPK and SREBP-1. J. Hepatol. 2009, 51, 535–547. [Google Scholar] [CrossRef] [PubMed]
  10. Azzalini, L.; Ferrer, E.; Ramalho, L.N.; Moreno, M.; Domínguez, M.; Colmenero, J.; Peinado, V.I.; Barberà, J.A.; Arroyo, V.; Ginès, P.; et al. Cigarette smoking exacerbates nonalcoholic fatty liver disease in obese rats. Hepatology 2010, 51, 1567–1576. [Google Scholar] [CrossRef]
  11. Tsuneto, A.; Hida, A.; Sera, N.; Imaizumi, M.; Ichimaru, S.; Nakashima, E.; Seto, S.; Maemura, K.; Akahoshi, M. Fatty liver incidence and predictive variables. Hypertens. Res. 2010, 33, 638–643. [Google Scholar] [CrossRef] [PubMed]
  12. Hamabe, A.; Uto, H.; Imamura, Y.; Kusano, K.; Mawatari, S.; Kumagai, K.; Kure, T.; Tamai, T.; Moriuchi, A.; Sakiyama, T.; et al. Impact of cigarette smoking on onset of nonalcoholic fatty liver disease over a 10-year period. J. Gastroenterol. 2011, 46, 769–778. [Google Scholar] [CrossRef]
  13. Koch, M.; Borggrefe, J.; Schlesinger, S.; Barbaresko, J.; Groth, G.; Jacobs, G.; Lieb, W.; Laudes, M.; Müller, M.J.; Bosy-Westphal, A.; et al. Association of a lifestyle index with MRI-determined liver fat content in a general population study. J. Epidemiol. Community Health 2015, 69, 732–737. [Google Scholar] [CrossRef] [PubMed]
  14. Suomela, E.; Oikonen, M.; Virtanen, J.; Parkkola, R.; Jokinen, E.; Laitinen, T.; Hutri-Kähönen, N.; Kähönen, M.; Lehtimäki, T.; Taittonen, L.; et al. Prevalence and determinants of fatty liver in normal-weight and overweight young adults. The Cardiovascular Risk in Young Finns Study. Ann. Med. 2015, 47, 40–46. [Google Scholar] [CrossRef] [PubMed]
  15. Zhang, T.; Zhang, C.; Zhang, Y.; Tang, F.; Li, H.; Zhang, Q.; Lin, H.; Wu, S.; Liu, Y.; Xue, F. Metabolic syndrome and its components as predictors of nonalcoholic fatty liver disease in a northern urban Han Chinese population: A prospective cohort study. Atherosclerosis 2015, 240, 144–148. [Google Scholar] [CrossRef] [PubMed]
  16. Kim, T.J.; Sinn, D.H.; Min, Y.W.; Son, H.J.; Kim, J.J.; Chang, Y.; Baek, S.Y.; Ahn, S.H.; Lee, H.; Ryu, S. A cohort study on Helicobacter pylori infection associated with nonalcoholic fatty liver disease. J. Gastroenterol. 2017, 52, 1201–1210. [Google Scholar] [CrossRef]
  17. Liu, P.; Xu, Y.; Tang, Y.; Du, M.; Yu, X.; Sun, J.; Xiao, L.; He, M.; Wei, S.; Yuan, J.; et al. Independent and joint effects of moderate alcohol consumption and smoking on the risks of non-alcoholic fatty liver disease in elderly Chinese men. PLoS ONE 2017, 12, e0181497. [Google Scholar] [CrossRef] [PubMed]
  18. van den Berg, E.H.; Amini, M.; Schreuder, T.C.; Dullaart, R.P.; Faber, K.N.; Alizadeh, B.Z.; Blokzijl, H. Prevalence and determinants of non-alcoholic fatty liver disease in lifelines: A large Dutch population cohort. PLoS ONE 2017, 12, e0171502. [Google Scholar] [CrossRef]
  19. Bayerl, C.; Lorbeer, R.; Heier, M.; Meisinger, C.; Rospleszcz, S.; Schafnitzel, A.; Patscheider, H.; Auweter, S.; Peters, A.; Ertl-Wagner, B.; et al. Alcohol consumption, but not smoking is associated with higher MR-derived liver fat in an asymptomatic study population. PLoS ONE 2018, 13, e0192448. [Google Scholar] [CrossRef]
  20. Okamoto, M.; Miyake, T.; Kitai, K.; Furukawa, S.; Yamamoto, S.; Senba, H.; Kanzaki, S.; Deguchi, A.; Koizumi, M.; Ishihara, T.; et al. Cigarette smoking is a risk factor for the onset of fatty liver disease in nondrinkers: A longitudinal cohort study. PLoS ONE 2018, 13, e0195147. [Google Scholar] [CrossRef]
  21. Okamura, T.; Hashimoto, Y.; Hamaguchi, M.; Obora, A.; Kojima, T.; Fukui, M. Low urine pH is a risk for non-alcoholic fatty liver disease: A population-based longitudinal study. Clin. Res. Hepatol. Gastroenterol. 2018, 42, 570–576. [Google Scholar] [CrossRef] [PubMed]
  22. Wang, L.; Li, M.; Zhao, Z.; Xu, M.; Lu, J.; Wang, T.; Chen, Y.; Wang, S.; Dai, M.; Hou, Y.; et al. Ideal Cardiovascular Health Is Inversely Associated with Nonalcoholic Fatty Liver Disease: A Prospective Analysis. Am. J. Med. 2018, 131, e1511–e1515. [Google Scholar] [CrossRef]
  23. Jung, H.S.; Chang, Y.; Kwon, M.J.; Sung, E.; Yun, K.E.; Cho, Y.K.; Shin, H.; Ryu, S. Smoking and the Risk of Non-Alcoholic Fatty Liver Disease: A Cohort Study. Am. J. Gastroenterol. 2019, 114, 453–463. [Google Scholar] [CrossRef] [PubMed]
  24. van den Berg, E.H.; Gruppen, E.G.; Blokzijl, H.; Bakker, S.J.L.; Dullaart, R.P.F. Higher Sodium Intake Assessed by 24 Hour Urinary Sodium Excretion Is Associated with Non-Alcoholic Fatty Liver Disease: The PREVEND Cohort Study. J. Clin. Med. 2019, 8, 2157. [Google Scholar] [CrossRef] [PubMed]
  25. Chen, X.; Ma, T.; Yip, R.; Perumalswami, P.V.; Branch, A.D.; Lewis, S.; Crane, M.; Yankelevitz, D.F.; Henschke, C.I. Elevated prevalence of moderate-to-severe hepatic steatosis in World Trade Center General Responder Cohort in a program of CT lung screening. Clin. Imaging 2020, 60, 237–243. [Google Scholar] [CrossRef] [PubMed]
  26. Okamura, T.; Hashimoto, Y.; Hamaguchi, M.; Obora, A.; Kojima, T.; Fukui, M. Creatinine-to-bodyweight ratio is a predictor of incident non-alcoholic fatty liver disease: A population-based longitudinal study. Hepatol. Res. 2020, 50, 57–66. [Google Scholar] [CrossRef]
  27. Takenaka, H.; Fujita, T.; Masuda, A.; Yano, Y.; Watanabe, A.; Kodama, Y. Non-Alcoholic Fatty Liver Disease Is Strongly Associated with Smoking Status and Is Improved by Smoking Cessation in Japanese Males: A Retrospective Study. Kobe J. Med. Sci. 2020, 66, E102–E112. [Google Scholar] [PubMed]
  28. Zhang, Q.; Ma, X.; Xing, J.; Shi, H.; Yang, R.; Jiao, Y.; Chen, S.; Wu, S.; Zhang, S.; Sun, X. Serum Uric Acid Is a Mediator of the Association Between Obesity and Incident Nonalcoholic Fatty Liver Disease: A Prospective Cohort Study. Front. Endocrinol. 2021, 12, 657856. [Google Scholar] [CrossRef] [PubMed]
  29. Jeong, S.; Oh, Y.H.; Choi, S.; Chang, J.; Kim, S.M.; Park, S.J.; Cho, Y.; Son, J.S.; Lee, G.; Park, S.M. Association of Change in Smoking Status and Subsequent Weight Change with Risk of Nonalcoholic Fatty Liver Disease. Gut Liver 2023, 17, 150–158. [Google Scholar] [CrossRef]
  30. Sadeghianpour, Z.; Cheraghian, B.; Farshchi, H.R.; Asadi-Lari, M. Non-alcoholic fatty liver disease and socioeconomic determinants in an Iranian cohort study. BMC Gastroenterol. 2023, 23, 350. [Google Scholar] [CrossRef]
  31. Akhavan Rezayat, A.; Dadgar Moghadam, M.; Ghasemi Nour, M.; Shirazinia, M.; Ghodsi, H.; Rouhbakhsh Zahmatkesh, M.R.; Tavakolizadeh Noghabi, M.; Hoseini, B.; Akhavan Rezayat, K. Association between smoking and non-alcoholic fatty liver disease: A systematic review and meta-analysis. SAGE Open Med. 2018, 6, 2050312117745223. [Google Scholar] [CrossRef]
  32. Higgins, J.P.; Thompson, S.G. Quantifying heterogeneity in a meta-analysis. Stat. Med. 2002, 21, 1539–1558. [Google Scholar] [CrossRef] [PubMed]
  33. Borenstein, M.; Hedges, L.V.; Higgins, J.P.; Rothstein, H.R. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res. Synth. Methods 2010, 1, 97–111. [Google Scholar] [CrossRef] [PubMed]
  34. Artese, A.; Stamford, B.A.; Moffatt, R.J. Cigarette Smoking: An Accessory to the Development of Insulin Resistance. Am. J. Lifestyle Med. 2017, 13, 602–605. [Google Scholar] [CrossRef]
  35. Cetin, E.G.; Demir, N.; Sen, I. The Relationship between Insulin Resistance and Liver Damage in non-alcoholic Fatty Liver Patients. Sisli Etfal Hastan. Tıp Bul. 2020, 54, 411–415. [Google Scholar] [CrossRef] [PubMed]
  36. Eliasson, B.; Attvall, S.; Taskinen, M.R.; Smith, U. The insulin resistance syndrome in smokers is related to smoking habits. Arterioscler. Thromb. 1994, 14, 1946–1950. [Google Scholar] [CrossRef]
  37. Utzschneider, K.M.; Kahn, S.E. Review: The role of insulin resistance in nonalcoholic fatty liver disease. J. Clin. Endocrinol. Metab. 2006, 91, 4753–4761. [Google Scholar] [CrossRef]
  38. Jia, W.P. The impact of cigarette smoking on metabolic syndrome. Biomed. Environ. Sci. 2013, 26, 947–952. [Google Scholar]
  39. Zhang, C.X.; Guo, L.K.; Qin, Y.M.; Li, G.Y. Association of polymorphisms of adiponectin gene promoter-11377C/G, glutathione peroxidase-1 gene C594T, and cigarette smoking in nonalcoholic fatty liver disease. J. Chin. Med. Assoc. 2016, 79, 195–204. [Google Scholar] [CrossRef] [PubMed]
  40. Myung, S.K. How to review and assess a systematic review and meta-analysis article: A methodological study (secondary publication). J. Educ. Eval. Health Prof. 2023, 20, 24. [Google Scholar] [CrossRef] [PubMed]
  41. Mackay, J.; Eriksen, M.P.; Shafey, O.; American Cancer Society. The Tobacco Atlas, 2nd ed.; American Cancer Society: Atlanta, GA, USA, 2006; p 1 atlas; 128p. [Google Scholar]
  42. Jung-Choi, K.H.; Khang, Y.H.; Cho, H.J. Hidden female smokers in Asia: A comparison of self-reported with cotinine-verified smoking prevalence rates in representative national data from an Asian population. Tob. Control 2012, 21, 536–542. [Google Scholar] [CrossRef]
  43. Hwang, J.-e.; Choi, Y.; Yang, Y.-s.; Oh, Y. Gender differences in the perceived effectiveness of female-focused graphic health warnings against smoking in South Korea. Health Educ. J. 2020, 79, 58–72. [Google Scholar] [CrossRef]
  44. Seo, D.C.; Torabi, M.R.; Kim, N.; Lee, C.G.; Choe, S. Smoking among East Asian college students: Prevalence and correlates. Am. J. Health Behav. 2013, 37, 199–207. [Google Scholar] [CrossRef] [PubMed]
  45. Nakhaee, N.; Divsalar, K.; Bahreinifar, S. Prevalence of and factors associated with cigarette smoking among university students: A study from Iran. Asia Pac. J. Public Health 2011, 23, 151–156. [Google Scholar] [CrossRef]
Gastroent 16 00001 g001
Figure 1. Flow diagram of identifying relevant Studies.
Figure 1. Flow diagram of identifying relevant Studies.
Gastroent 16 00001 g001
Gastroent 16 00001 g002
Figure 2. Smoking and risk of fatty liver disease in the meta-analysis (n = 20). OR, odds ratio; RR, relative risk; HR, hazard ratio; CI, confidence interval.
Figure 2. Smoking and risk of fatty liver disease in the meta-analysis (n = 20). OR, odds ratio; RR, relative risk; HR, hazard ratio; CI, confidence interval.
Gastroent 16 00001 g002
Gastroent 16 00001 g003
Figure 3. Begg’s funnel plot and Egger’s test to test publication bias (n = 20). OR, odds ratio; RR, relative risk; HR, hazard ratio; S.E., standard error.
Figure 3. Begg’s funnel plot and Egger’s test to test publication bias (n = 20). OR, odds ratio; RR, relative risk; HR, hazard ratio; S.E., standard error.
Gastroent 16 00001 g003
Table 1. General characteristics of the studies Included in the final analysis (n = 20).
Table 1. General characteristics of the studies Included in the final analysis (n = 20).
Study RegionType of StudyGenderStudy Participants (% of Men)ComparisonOdds Ratio, Relative Risk, or Hazard Ratio, and 95% Confidence IntervalOutcomes Adjusted Variables
2010 Tsuneto [11]AsiaProspectiveBothA total of 1635 atomic bomb
survivors who underwent biennial examinations in Nagasaki without NAFLD at baseline		Ex-smoker or current smoker vs. none0.92 (0.64–13.4)FLDAge, sex, BMI, DM, HTN, dyslipidemia, drinking habits, and atomic radiation dose
2011 Hamabe [12]AsiaRetrospectiveBothA total of 1560 subjects without NAFLD who underwent a complete medical health checkup at the Kagoshima Kouseiren Medical Healthcare CenterCigarette smoking vs. no smoking1.44 (0.86–2.42)NAFLDAge, sex, obesity, HTN, dyslipidemia, dysglycemia, and alcohol intake
2015 Koch [13]EuropeCross-sectionalBothA total of 747 official population registeries in KielCigarette smoking vs. no smoking1.14 (0.71–1.82)FLDAge, sex, years of education, total energy intake, physical activity, and waist circumference
2015 Suomela [14]EuropeCross-sectionalBothA total of 3592 Young FinnsCurrent smoker vs. none2.56 (1.18–5.52)FLDAge, sex, BMI, and waist circumference
2015 Zhang [15]AsiaProspectiveBothA total of 15,791 health check-up participants at the Center for Health Management of Shandong Provincial
Qianfoshan Hospital and Shandong Provincial Hospital		Current smoker vs. none1.03 (0.95–1.11)NAFLDBaseline Mets status, sex, age, diet, smoking status, and regular exercise
2017 Kim [16]AsiaRetrospectiveBothA total of 17,028 health-screening exam participants at the Center for Health Promotion of the Samsung Medical Center, South KoreaCurrent smoker vs. none0.96 (0.81–1.15)NAFLDAge, sex, body mass index, year of screening exam, alcohol
intake, regular exercise, and education level
2017 Liu [17]AsiaCross-sectionalMaleA total of 9432 DFTJ cohort study among retirees of Dong feng Motor corporationCurrent smoker vs. none1.52 (1.22–1.88)NAFLDAge, body mass index, waist circumference alcohol
intake, DM, HTN, dyslipidemia, and past history of CHD
2017 van den Berg [18]EuropeCross-sectionalBothA total of 37,496 participants in the Framework of the Lifelines Cohort StudyCurrent smoker vs. none1.32 (1.21–1.43)FLDAge, sex, Hemoglobin, ALT/ALP/Albumin, HBA1c, type 2 DM, dyslipidemia, and past history of CHD
2018 Bayerl [19]EuropeCross-sectionalBothA total of 1282 persons from Cooperative Health Research in the German regionEx-smoker or current smoker vs. none0.56 (0.27–1.17)FLDAge, sex, DM, and alcohol intake
2018 Okamoto [20]AsiaRetrospectiveBothA total of 7905 persons who underwent a health checkup at the Ehime General Healthcare AssociationCurrent smoker vs. none2.25 (1.10–4.38)FLDAge, sex, BMI, DM, HTN, CVD, dyslipidemia, and snacking habit
2018 Okamura [21]AsiaRetrospectiveBothA total of 29,555 people in the medical
examination program at the Murakami Memorial Hospital using the NAGALA
(NAFLD in the Gifu Area, Longitudinal Analysis) database		Current smoker vs. none0.88 (0.78–0.99)NAFLDAge, sex, BMI, ALT, triglycerides, exercise habit, alcohol consumption, systolic blood pressure, fasting
plasma glucose, and uric acid
2018 Wang [22]AsiaProspectiveBothA total of 10,375 participants from community residents in the Jiading District of ShangiaCurrent smoker or quit < 12 mo vs. none1.11 (0.78–1.56)NAFLDAge, sex, alcohol consumption, education, and HOMA-IR
2019 Jung [23]AsiaProspectiveBothA total of 199,468 persons who underwent a health checkup by the Kangbuk Samsung Health StudyEx-smoker or current smoker vs. noneMen: 1.15
(1.12–1.18)
Women: 1.14
(1.03–1.27)		FLDAge, sex, BMI, DM, HTN, dyslipidemia, alcohol drinking, education level, physical activity, waist circumference, and laboratory test
2019 van den Berg [24]EuropeCross-sectionalBothA total of 6132 participants in the prevention of Renal and Vascular End-stage Disease cohort studyCurrent smoker vs. none1.24 (1.05–1.46)NAFLDAge, sex, BMI, DM, HTN, dyslipidemia, alcohol drinking, estimated GFR, urine albumin excretion, use of antihypertensive medication, glucose lowering drugs, lipid lowering drugs, and HOMA-IR
2020 Chen [25]USCross-sectionalBothA total of 154 World Trade Center participants in NIOSHCurrent smoker vs. none0.41 (0.17–0.99)FLDAge, sex, Ethnicity, BMI, DM, HTN, COPD, and membership in the WTC
2020 Okamura [26]AsiaRetrospectiveBothA total of 13,728 population-based longitudinal study
of participants in a medical checkup program at Asahi
University Hospital		Current smoker vs. none1.16 (0.88–1.52)NAFLDAge, aspartate aminotransferase, fasting plasma glucose, triglyceride to high-density lipoprotein cholesterol ratio, systolic blood pressure, alcohol consumption, and exercise.
2020 Takenaka [27]AsiaCross-sectionalBothA total of 8297 health check-up participants at
Yodogawa Christian Hospital		Current smoker vs. none1.31 (1.17–1.47)NAFLDAge, sex, presence of metabolic syndrome, and light alcohol consumption
2021 Zhang [28]AsiaProspectiveBothA total of 16,839 participants who received the Kailuan Group’s detailed
and thorough medical examination at Tangshan City, China		Current smoker vs. none1.15 (1.06–1.25)NAFLDAge, sex, marital status, working type, education level, physical activity, systolic blood pressure, lipid profile, CRP, and Cr
2023 Jeong [29]AsiaRetrospectiveBothA total of 296,033 in the NHIS of KoreaCurrent smoker vs. none1.64 (1.39–1.94)FLDAge, sex, household income, BMI, HTN, DM, HL, physical activity, and Charlson comorbidity index
2023 Sadeghianpour [30]AsiaCross-sectionalBothA total of 180,000 Iranian adults in a Hoveyzeh Cohort Study Current smoker vs. none0.63 (0.50–0.79)FLDAge, sex, area, physical activity, Energy intake household income, DM, HL, education level, wealth status, and skill level
Abbreviations: NAFLD, non-alcoholic fatty liver disease; FLD, fatty liver disease; DMC, Dong feng Motor Corporation; BMI, body mass index; DM, Diabetes Mellitus; HTN, Hypertension; HL, Hyperlipidemia; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; NIOSH, National Institute of Occupational Safety and Health; NHIS, National Health Insurance Service; HOMA-IR, homeostasis model assessment of insulin resistance; GFR, glomerular filtration rate.
Table 2. Methodological quality of studies assessed by the Newcastle–Ottawa Scale (n = 11) *.
Table 2. Methodological quality of studies assessed by the Newcastle–Ottawa Scale (n = 11) *.
StudiesSelectionComparabilityExposureTotal Score
12341123
Representativeness of the Exposed CohortSelection of the Non-Exposed CohortAscertainment of ExposureOutcome of Interest Not Present at Start of the StudyComparability of CohortsAssessment of OutcomeAdequate Follow-Up Period for Outcome of Interest Adequacy of Follow-Up of Cohorts
2010 Tsuneto [11]011121107
2011 Hamabe [12]110121107
2015 Zhang [15]110121118
2017 Kim [16]110121118
2018 Okamoto [20]110121118
2018 Okamura [21]110121107
2018 Wang [22]111121108
2019 Jung [23]110121107
2020 Okamura [26]110121118
2021 Zhang [28]110121107
2023 Jeong [29]110121118
* Among the 20 cohort studies included in the final analysis, cross-sectional surveys were excluded for the assessment of methodological quality. The average score of all the prospective and retrospective cohort studies is 7.5.
Table 3. Association between smoking and fatty liver disease in subgroup meta-analysis by various factors.
Table 3. Association between smoking and fatty liver disease in subgroup meta-analysis by various factors.
FactorNo. of StudiesRR (95% CI)Heterogeneity I2 (%)
All Studies201.14 (1.05–1.24) *83.7
Type of cohort study
   Prospective51.15 (1.05–1.18) *51.7
   Retrospective61.23 (0.94–1.62)88.4
   Cross-sectional91.12 (0.92–1.36)85.2
Region
   Europe81.32 (1.16–1.50) *82.1
   Asia101.03 (0.91–1.18)83.5
   US20.75 (0.28–2.06)79.5
Type of fatty liver disease
   Fatty liver disease61.27 (1.01–1.59) *72.7
   Non-alcoholic fatty liver disease141.09 (1.00–1.19) *83.2
Gender (All from Asia)
   Men 41.15 (1.06–1.25) *71.2
   Women 41.12 (0.94–1.34)47.8
Follow-up period
   <5 years31.44 (0.95–2.13)93.0
   >5 years71.08 (0.98–1.19)70.9
Quality of study †
   High61.16 (0.94–1.42)85.3
   Low51.07 (0.95–1.20)67.1
* Indicates a significant association. RR, relative risk; CI, confident interval. † Study quality was assessed based on the Newcastle–Ottawa Scale.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lee, M.; Myung, S.-K.; Lee, S.H.; Chang, Y. Smoking and Risk of Fatty Liver Disease: A Meta-Analysis of Cohort Studies. Gastroenterol. Insights 2025, 16, 1. https://doi.org/10.3390/gastroent16010001

AMA Style

Lee M, Myung S-K, Lee SH, Chang Y. Smoking and Risk of Fatty Liver Disease: A Meta-Analysis of Cohort Studies. Gastroenterology Insights. 2025; 16(1):1. https://doi.org/10.3390/gastroent16010001

Chicago/Turabian Style

Lee, Moonhyung, Seung-Kwon Myung, Sang Hee Lee, and Yoosoo Chang. 2025. "Smoking and Risk of Fatty Liver Disease: A Meta-Analysis of Cohort Studies" Gastroenterology Insights 16, no. 1: 1. https://doi.org/10.3390/gastroent16010001

APA Style

Lee, M., Myung, S.-K., Lee, S. H., & Chang, Y. (2025). Smoking and Risk of Fatty Liver Disease: A Meta-Analysis of Cohort Studies. Gastroenterology Insights, 16(1), 1. https://doi.org/10.3390/gastroent16010001

Article Metrics

Back to TopTop