Article Text

Download PDFPDF

Original research
Epidemiology of chronic liver diseases in the USA in the past three decades
  1. Zobair M Younossi1,2,
  2. Maria Stepanova3,
  3. Youssef Younossi3,
  4. Pegah Golabi1,2,
  5. Alita Mishra1,2,
  6. Nila Rafiq1,2,
  7. Linda Henry3
  1. 1 Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, Virginia, USA
  2. 2 Department of Medicine, Center For Liver Disease, Inova Fairfax Hospital, Falls Church, Virginia, USA
  3. 3 Department of Statistical Services, Center for Outcomes Research in Liver Diseases, Washington, DC, USA
  1. Correspondence to Dr Zobair M Younossi, Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA 22042, USA; zobair.younossi{at}


Objective Given significant advances in treatment of viral hepatitis and the growing epidemic of obesity, the burden of the different types of liver diseases in the USA may be changing. Our aim was to assess the shift in the prevalence of different liver disease aetiologies in the USA over the past three decades.

Design National Health and Nutrition Examination Surveys (NHANES; cross-sectional 1988–1994 and 1999–2016) were used.

Results A total of 58 731 adults from NHANES (1988–2016) were included. Over the study period, the prevalence of chronic hepatitis B and alcoholic liver disease remained stable: 0.3%–0.4% and 0.8%–1.0%, respectively (p>0.05). The prevalence of chronic hepatitis C decreased nearly twofold: 1.6% in 1988–1994 to 0.9% in 2013–2016 (p=0.03). In contrast, the prevalence of non-alcoholic fatty liver disease (NAFLD; by US-Fatty Liver Index) increased from 20.0% (1988–1994) to 28.3% (1999–2004) to 33.2% (2009–2012) and 31.9% (2013–2016) (p<0.0001). Furthermore, steady increases were observed in the rates of obesity (22.2% in 1988–1994 to 31.0% in 1999–2004 to 38.9% in 2013–2016), type 2 diabetes mellitus (T2DM) (from 7.2% to 8.2% to 13.5% same years), insulin resistance and hypertension (all p<0.0001). Yearly trend analyses showed that the only LD with consistently increasing prevalence was NAFLD (trend p=0.01). Multivariable regression analysis showed that obesity (OR 10.4; 95% CI 9.5 to 11.3) and T2DM (OR 3.7; 95% CI 3.2 to 4.2) were the major independent predictors of NAFLD.

Conclusions Over the past 30 years in the USA, NAFLD is the only liver disease with growing prevalence, synchronous with the increasing rates of obesity and T2DM.

  • non-alcoholic fatty liver disease
  • epidemiology
  • trends
  • obesity
  • diabetes

Statistics from

Significance of this study

What is already known on this subject?

  • Liver disease is a major cause of morbidity and mortality with 1.75 million individuals dying of liver disease each year worldwide.

  • Although chronic viral hepatitis B and hepatitis C infections are the two major causes, alcohol-related liver disease and non-alcoholic fatty liver disease (NAFLD) have been increasingly contributing to the burden of liver disease.

What are the new findings?

  • Over the study period, the prevalence of hepatitis B and alcohol-related liver disease remained stable.

  • The prevalence of hepatitis C decreased nearly twofold.

  • At the same time, the prevalence of NAFLD increased substantially.

How might it impact on clinical practice in the foreseeable future?

  • Prevention of major NAFLD drivers which are obesity and diabetes may help to decreased the disease burden.


Among aetiologies of chronic liver disease (CLD), chronic hepatitis B (CHB), HCV, alcohol-related liver disease (ALD) and non-alcoholic fatty liver disease (NAFLD) are the most important contributors to the global burden of liver disease.1 However, over the past decade, development of new antiviral regimens and the global epidemic of obesity are potentially changing the face of CLD across the world. In this context, HCV can be cured with highly effective antiviral treatment regimens, and HBV can be prevented with vaccines and can be suppressed effectively with oral antiviral regimens in the chronic stage.2 As a result, the burden of viral hepatitis in the Western world seems to be decreasing although it still remains a challenge in the less developed parts of the world.1 2

On the other hand, the global epidemic of obesity has reached every corner of the world with large proportions of the general population being overweight or obese. In fact, in the USA, 60% of adults are considered overweight or obese.3 4 Obesity-associated diseases such as type 2 diabetes mellitus (T2DM) have also reached all-time high prevalence in most developed countries.5 In this context, liver diseases associated with obesity and T2DM, namely, NAFLD and non-alcoholic steatohepatitis (NASH), and their progression to cirrhosis and liver-related mortality seem to pose a growing global burden.6–19 In fact, although the rate of progression of NASH may be slower than that of HCV infection,7 8 the affected cohort of NASH is substantially larger leading to a greater number of patients with NASH presenting with cirrhosis and liver cancer.9 10 As a result, NASH is now considered the second most common indication for liver transplantation in the USA.11 Furthermore, NAFLD and NASH are among the top causes of hepatocellular carcinoma (HCC) in the USA.11–14 In fact, in a recent study of liver transplant candidates listed for HCC, NASH showed the steepest increase as an indication for HCC transplantation over the recent decades.13 Finally, NAFLD and NASH have been associated with a tremendous economic and health-related quality of life burden.15

In this study, we intend to assess recent changes in the prevalence of different liver diseases over the past three decades in the USA.


This study used the National Health and Nutrition Examination Survey (NHANES), a nationwide survey representing the health and nutritional status of the non-institutionalised civilian US population. The publicly available data were collected by the US National Center for Health Statistics of the Centers for Disease Control and Prevention via household interviews, physical examinations and laboratory data including blood and urine samples collected in designated examination centres.20 With accounting for the survey design (stratification and sampling) and selection probabilities (via population-based weights21), the distribution of adult participants is supposed to be representative of the US population.22 For the purpose of this study, we used the data from NHANES III (1988–1994) and from nine biannual cycles published so far (1999–2016).

In this study, we included NHANES participants of at least 20 years of age with clinical and laboratory data sufficient to rule in or rule out the presence of CLD (figure 1). Similarly to our older study on a similar topic,16 four major categories of CLD were considered: CHB (positive HBV surface antigen), chronic hepatitis C (CHC; positive HCV RNA), ALD (significant alcohol use in the presence of elevated liver enzymes, defined below) and NAFLD (the US-Fatty Liver Index (US-FLI) score ≥3023 in the absence of significant alcohol use and other CLD causes). Elevated serum aminotransferases were defined as alanine aminotransferase (ALT) >40 U/L or aspartate aminotransferase (AST) >37 U/L in men and ALT or AST >31 U/L in women. Significant alcohol use was defined as >30 g/day for men and >20 g/day for women24 within a year prior to data collection, based on self-reported data on the amount and frequency of alcohol use.

Figure 1

Flowchart for the study sample. NHANES, National Health and Nutrition Examination Survey; US-FLI, US-Fatty Liver Index.

Other definitions used in this study were also similar to those from our previous study.16 Namely, T2DM was defined as a fasting glucose ≥126 mg/dL or self-reported history or the use of oral hypoglycaemics or insulin. Hypertension was defined as clinical history or systolic blood pressure (BP) of 140 mm Hg or greater, diastolic BP of 90 mm Hg or greater or the use of oral antihypertensive medications. Hypercholesterolaemia was defined as an elevated cholesterol of >200 mg/dL, low-density lipoprotein ≥139 mg/dL or high-density lipoprotein (HDL) <40 mg/dL for men and <50 mg/dL for women. Insulin resistance (IR) was determined by the homeostasis model assessment score (HOMA).25 A HOMA score of 3.0 or greater was defined as IR. Obesity was defined as body mass index (BMI) ≥30.

Statistical analyses

The data were merged into five study cycles: 1988–1994, 1999–2004, 2005–2008, 2009–2012 and 2013–2016; two most recent cycles represent newly added data in comparison with our prior study. The prevalence of demographic and clinical parameters and the CLD aetiologies was calculated while accounting for the survey design and sampling weights and was compared across the five study cycles by using the stratum-specific χ2 test for independence. The time trends were assessed using Kendall correlation metric. Survey logistic regression was used to test independent association with the presence of NAFLD for clinical and demographic factors and also the study cycle. All analyses were run with SAS V.9.4.


This study included a cohort of 58 731 adults from the five NHANES cycles with clinical, demographic and laboratory data (figure 1). Over time, the US population is ageing, becomes more obese, more people have T2DM, IR and hypertension (all p<0.0001). In fact, in the most recent NHANES cycle, the prevalence of obesity approached 40% and that of T2DM was 13.5%, both more than 75% higher than the respective rates seen in 1988–1994 (table 1).

Table 1

Demographic characteristics and chronic liver disease prevalence in the USA in 1988–2016 (National Health and Nutrition Examination Survey)

Over the study period, the prevalence of CHB, CHC and ALD did not significantly increase (all one-sided p>0.10; table 1). In fact, the prevalence of CHC followed a significantly decreasing trend with the prevalence in 2013–2016; being 40% lower in 2013–2016 in comparison with the rate reported for 1988–1994 (p=0.03). In contrast, NAFLD has significantly increased overtime from a prevalence of 20% in 1988–1994 to 31.9% in 2013–2016 (p<0.0001; table 1). In fact, the trend analysis determined that the prevalence rate of NAFLD had an annual growth of +0.47 percentage points per year (p=0.0014) which was significantly higher in comparison with all other CLD aetiologies (p<0.0001) (figure 2).

Figure 2

Relative changes in the prevalence of chronic liver disease aetiologies (reference: 1988–1994 cycle). NAFLD, non-alcoholic fatty liver disease.

Independent predictors of NAFLD

Since NAFLD was the only CLD that was found to be increasing, we chose to perform further analysis to determine factors that are independently associated with NAFLD. Our multivariable regression analysis showed that after adjustment for age, gender, ethnicity and other clinical parameters, the strongest independent predictors of NAFLD were obesity (OR 10.4; 95% CI 9.5 to 11.3) and T2DM (OR 3.7; 95% CI 3.2 to 4.2) (p<0.0001) (table 2). Indeed, of those who have both obesity and T2DM, 80.5% also have NAFLD, as opposed to 10.9% in those who have neither obesity nor T2DM, and 47%–55% in those who have only one of the two conditions. Furthermore, after adjustment for only these two factors, without accounting for other NAFLD predictors, the association of the study cycle with the risk of NAFLD is no longer significant (OR 1.13; 95% CI 1.10 to 1.17 per study cycle, p<0.0001, without adjustment vs OR 1.03; 95% CI 0.99 to 1.06 per study cycle, p=0.14, with adjustment for only obesity and T2DM). This suggests that the increasing trends in T2DM and obesity could partly explain out the reported growth in the prevalence of NAFLD in the US population.

Table 2

Factors independently associated with having NAFLD in a multivariable logistic regression model.


Using three decades of NHANES data, this study reports that the prevalence of HBV and ALD in the USA remains flat, while the prevalence of HCV is beginning to decline. In contrast, the prevalence of NAFLD during the same period of time keeps increasing.

Although a steady decline in HCV was noted before 2010, the trend may get less clear in the more recent years in light of recent reports which suggest an increase in the incidence of HCV in the USA starting in early 2010s.26 The latter could be partially related to shifting demographics of the infection in young, non-Hispanic whites.27 However, given limited time span of our study, we were not able to quantify the effect of this on chronic HCV prevalence rates. In contrast, with the significant increases in the rates of obesity and T2DM, the rate of NAFLD was expected to grow. Indeed, we have shown that NAFLD has increased in prevalence alongside increases in obesity and diabetes and, furthermore, increases in prevalence of obesity and diabetes are likely to have contributed to the increase in prevalence of NAFLD. Nevertheless, we have also shown that this growth in NAFLD may have slowed down starting around early 2010s. That could potentially be explained by the deceleration of obesity rates around that time which, however, seems to have been only temporary.28 Notably, we have also found that Hispanic ethnicity was independently associated with an increased risk of NAFLD (46.0% vs 29.5% as of 2013–2016, adjusted OR of 2.7 with reference to white Americans across all study cycles; p<0.0001) which might be related to higher prevalence of unfavourable PNPLA3 genotype and, thus, greater susceptibility to NAFLD among individuals of Mexican origin.29

These data are in agreement with reports published from NHANES almost a decade ago.16 Additionally, the data are consistent with reports from the USA suggesting growing number of liver deaths attributed to NAFLD.17 In fact, in the recent years, we are witnessing more patients with NAFLD with advanced liver disease who are listed for liver transplantation and more patients with NAFLD who present with HCC.11–14 All of these data contribute to the mounting evidence for an increasing clinical burden of NAFLD in the USA. In fact, modelling studies have reported that the complications due to NASH are expected to double by 2030,10 while other similar studies have suggested that the direct lifetime costs of patients with advanced NASH in the USA are close to US$95 billion.18 Furthermore, in addition to the clinical and economic burden, patients with NAFLD have significant impairment of health-related quality of life and other patient-reported outcomes.13 19

Despite this increasing burden, diagnosis and treatment options for NAFLD and NASH remain limited.30 31 Currently, the most commonly used non-invasive tool to establish the diagnosis of NAFLD is ultrasound, while liver biopsy remains the most accurate method to diagnose NASH and to assess the stage of hepatic fibrosis.31 As a result, screening and diagnosis of NAFLD and NASH in the general population remains a challenging problem. Nevertheless, a great deal of effort around the world has focused on the development of non-invasive tests. Once validated, these tests can be used as a part of algorithms to screen patients with progressive type of NAFLD and link them to appropriate clinical management. It is important to note that the only recommended treatments for NAFLD at present are diet and exercise; however, the exact dietary regimen or the type of exercise (weightlifting, aerobic or both), its duration and intensity have not been established.

Over the next few years, it is expected that non-invasive tests to establish stage of fibrosis in NASH will become available. Furthermore, targeted drug regimens may provide additional options for patients with advanced NASH. Nevertheless, the data from this analysis show that the burden of NAFLD and NASH will not be addressed without addressing the most likely root of the problem which is obesity and T2DM. In fact, according to recent national reports, up to one-quarter of T2DM cases remain undiagnosed.32 In this context, policymakers, providers and patients must focus on public health programmes that can address the growing burden of obesity and T2DM. These efforts should be coupled with studies to develop drug regimens for patients with advanced NASH. Additionally, awareness about T2DM and NASH and their complications must increase among all stakeholders, including patients, providers and policymakers via both medical and patients’ education.33–36

A limitation of our study is the use of US-FLI for the diagnosis of NAFLD. Although this non-invasive score has been validated and is reported to be an accurate measure to use when determining the presence of NAFLD, our data still lacks histological confirmation.23 37 38 In addition, we identified patients with NAFLD using the common exclusion-based definition while hepatic steatosis can, in fact, be superimposed on other causes of CLD, such as ALD or viral hepatitis, and both could contribute to the disease progression as reported in a recent study using the same data set.39 Nevertheless, given the large number of subjects derived from a population-based survey with three decades of data, our analysis provides strong evidence regarding the changes in the burden of different aetiologies of CLD in the USA.

In conclusion, over the past 30 years, the portrait of CLDs in the USA has changed. While hepatitis B and alcohol-related liver disease remained stable, the prevalence of hepatitis C seems to be decreasing. In contrast, the prevalence of NAFLD in the USA is increasing alongside the epidemic of obesity and T2DM. Since NAFLD and its progressive form of NASH are associated with significant clinical, economic and quality of life burden, a multiprong approach to deal with NASH must be undertaken. These efforts must include public health programmes, development of accurate non-invasive diagnostic tools and targeted treatment regimens for NASH. It is only with this comprehensive approach that the global epidemic of NAFLD and NASH can be addressed.


Supplementary materials

Related Data


  • Contributors ZMY is the guarantor of the article and contributed to the study concept and design; obtained funding and supervised the study. MS contributed to the acquisition of data, analysis and interpretation of data and statistical analysis. YY and PG gave technical or material support. AM and NR contributed to the study concept and design and interpretation of data. All authors contributed to the critical revision of the manuscript for important intellectual content.

  • Funding Internal funds only.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Ethics approval The study was approved by Inova Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data may be obtained from a third party and are not publicly available.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.