What Do We Know about Inequalities in NAFLD Distribution and Outcomes? A Scoping Review

With prevalence high and rising given the close relationship with obesity and diabetes mellitus, non-alcoholic fatty liver disease (NAFLD) is progressively becoming the most common chronic liver condition worldwide. However, little is known about the health inequalities in NAFLD distribution and outcomes. This review aims to analyze health inequalities in NAFLD distribution globally and to assess the health disparities in NAFLD-related outcomes. We conducted a scoping review of global health inequalities in NAFLD distribution and outcomes according to gender/sex, ethnicity/race, and socioeconomic position from PubMed’s inception to May 2021. Ultimately, 20 articles were included in the review, most (75%) of them carried out in the United States. Males were found to have a higher NAFLD prevalence (three articles), while available evidence suggests that women have an overall higher burden of advanced liver disease and complications (four articles), whereas they are less likely to be liver-transplanted once cirrhosis develops (one article). In the US, the Hispanic population had the highest NAFLD prevalence and poorer outcomes (seven articles), whereas Whites had fewer complications than other ethnicities (two articles). Patients with low socioeconomic status had higher NAFLD prevalence (four articles) and a higher likelihood of progression and complications (five articles). In conclusion, globally there is a lack of studies analyzing NAFLD prevalence and outcomes according to various axes of inequality through joint intersectional appraisals, and most studies included in our review were based on the US population. Available evidence suggests that NAFLD distribution and outcomes show large inequalities by social group. Further research on this issue is warranted.


Introduction
Non-alcoholic fatty liver disease (NAFLD) encompasses a spectrum of liver diseases ranging from simple hepatic steatosis (fatty liver) to steatohepatitis (NASH) to cirrhosis and hepatocellular carcinoma (HCC) [1]. NAFLD is currently considered the liver manifestation of metabolic syndrome (which also encompasses obesity, type 2 diabetes mellitus (T2DM), insulin resistance, and dyslipidemia) [2]. A significant part of the scientific community

Selection of the Review Approach
A scoping review approach was chosen because this topic has not been extensively reviewed to date and has a complex and heterogeneous nature [16]. The methodologic proposal of Arksey and O'Malley was followed [17]. PRISMA guidelines were followed [18].

Literature Search Strategy
A PubMed search was conducted using the following keywords and expressions: ["NAFLD" OR "NASH" OR "fatty liver" OR "MAFLD") AND/OR ("prevalence" AND/OR "distribution" AND/OR "incidence", AND/OR "epidemiology") AND/OR ["cirrhosis" AND/OR "hepatocellular carcinoma" OR "liver cancer", AND/OR "advanced fibrosis" AND/OR "cardiovascular events" AND/OR "hepatic events", AND/OR "mortality"] AND ("gender" AND/OR "sex" AND/OR "ethnicity" AND/OR "race" AND/OR "universal health coverage" AND/OR "socioeconomic position" AND/OR "socioeconomic status" AND/OR "education" AND/OR "housing", AND/OR "income") AND/OR ["disparities" OR "inequities" OR "equity" OR "inequalities"] from PubMed inception to 31 May 2021. Furthermore, a manual search of secondary references from screened articles was performed.

Eligibility Criteria
Articles were included if all the following inclusion criteria were met: (i) articles written in English; (ii) original articles; (iii) studies of at least one element of each of the two following pairs: NAFLD or MAFLD, distribution or outcomes (see definitions below); and (iv) studies analyzing at least one axis of inequality (e.g., articles including SDH as subvariables for analysis without specific analysis of how they impact disparities in NAFLD distribution or outcomes were not included).
Articles were excluded if they: (i) mostly analyzed other causes of hepatic steatosis such as chronic viral hepatitis (C and/or B hepatitis viruses), drugs, or alcohol-related liver disease; (ii) did not exclude patients with elevated alcohol consumption (over 20 g/day in females and over 30 g/day in males), even if the most likely diagnosis was NAFLD; (iii) focused on pediatric population (<18 years); (iv) the study sample was smaller than 150 individuals; and (v) were reviews, meta-analyses, editorials, or commentaries.

Study Selection and Data Extraction
Articles were first screened based on the search keywords with title and abstract. Subsequently, a screening applying the inclusion/exclusion criteria was conducted, and ultimately a full-text reading was performed.
Extracted data included: (i) first author name, (ii) publication year, (iii) period of study, (iv) country of study, (v) study design and population, (vi) method of diagnosis, (vii) distribution index/analyzed outcomes, (viii) axes of inequality, (ix) results, (x) conclusions, and (xi) limitations of the study. Finally, data were summarized in two tables depending on whether the study analyzed NAFLD distribution (Table 1) or analyzed other outcomes such as advanced liver disease or complications ( Table 2).

Definitions
The definitions that we used are as follows: Advanced fibrosis: Liver fibrosis in stages 3 and 4 [5]. Advanced liver disease: NAFLD/NASH with advanced fibrosis or cirrhosis. Axes of inequality: social determinants, such as social class, socioeconomic position, gender/sex, and ethnicity/race among, others that determine the opportunities for good health and reveal the existence of health inequalities due to hierarchies of power, prestige, or access to resources [19].
Cirrhosis: histological development of regenerative nodules surrounded by fibrous bands in response to chronic liver injury, which leads to portal hypertension and end-stage liver disease [20].
Complications of NAFLD: unfavorable results of the disease, comprising liver decompensation, liver cancer, liver transplant, cardiovascular events, hospitalization, and mortality.
Distribution of NAFLD: frequency and pattern of health events in the population, comprising prevalence and incidence.
Ethnicity/Race: historical social groups that share specific social, cultural, and some biological attributes [21].
Gender: a social construct regarding culture-bound conventions, roles, and behaviors for, as well as relationships between and among women and men and boys and girls, as opposed to sex, which is determined by the genetic background and secondary sexual characteristics [21].
Health inequalities: differences in health status or in the distribution of health resources between different population groups, which are not only unnecessary and avoidable, but in addition, are considered unfair and unjust [22].
Liver decompensation: acute deterioration in liver function in a patient with cirrhosis and is characterized by jaundice, ascites, hepatic encephalopathy, hepatorenal syndrome, or variceal hemorrhage [24].
Outcomes of NAFLD: NAFLD-related consequences comprising NAFLD progression to advanced liver disease and NAFLD-related complications.
Social determinants of health: conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life. These forces and systems include economic policies and systems, development agendas, social norms, social policies, and political systems [25].
Social groups: sets of people whose ways of living have been historically determined by one or various social determinants of health such as gender, ethnicity, migration status, or socioeconomic position [25].
Social inequalities: in the context of health, health disparities, within and between countries, that are judged to be unfair, unjust, avoidable, and unnecessary (meaning: are neither inevitable nor irremediable) and that systematically burden populations rendered vulnerable by underlying social structures and political, economic, and legal institutions [21].
Socioeconomic position/status: an aggregate concept that includes both resourcebased and prestige-based measures, as linked to both childhood and adult social class position. Resource-based measures refer to material and social resources and assets, including income, wealth, and educational credentials [21].

Literature Search
Of 603 articles retrieved from the initial search, 481 (79.8%) were excluded by title, and 122 (20.2%) articles were selected for abstract screening. Of 122 articles, 81 (66.4%) were excluded applying the inclusion/exclusion criteria, and 41 (33.6%) were selected for full-text reading. At this point, 26 studies were added from reference lists of other selected articles. After the full-text reading, 47 did not meet the inclusion/exclusion criteria. Therefore, a total of 20 articles were selected to perform the review (Figure 1).

Figure 1.
Flowchart of studies disposition throughout the search and article selection process according to PRISMA guidelines for scoping reviews [18].
NAFLD is related to the presence of AAC on Blacks and women. These suggest disparities in the pathophysiologic pathways in which atherosclerosis develops.
African Americans with NAFLD were significantly older than other racial or ethnic groups (p < 0.001), and in Asians, NAFLD was 3.5 times more common in males than in females (p = 0.016). Hispanics with NAFLD were over-represented (28-10%) and whites were under-represented (45-59%).
These racial and gender variations may reflect differences in genetic susceptibility to visceral adiposity, including hepatic involvement, and may have implications for the evaluation of persons with metabolic syndrome. The prevalence of food insecurity in patients with NAFLD is significantly higher compared to controls.

Gender/Sex
In the general population, NAFLD is more prevalent among men than women [26][27][28]. Specifically, in China, there is an NAFLD prevalence of 26.5% in men and 20.1% in women [27], and among US citizens, men comprise 58% of the total population with NAFLD [26]. However, this prevalence disparity is not conserved when models are adjusted for body mass index and age. Among lean or non-overweight populations, women seem to have a higher prevalence than men [23], while when only postmenopausal women are taken into account, NAFLD prevalence seems to be the same between males and females [28,29]. Additionally, men tend to have an NAFLD onset sooner than women. Younger men (≤45 years) have a higher prevalence than older men (30% vs. 24%) [27]. Conversely, women have a higher prevalence at an older age. In the Chinese population, women aged less than 45 years have a lower prevalence of NAFLD compared to women aged more than 45 years (15% vs. 22.8%) [27], and in the US population, elderly females are more likely to have NAFLD than younger females (28-14%) [37]. Further, it has also been described that increased age is associated with a decreased risk of having NAFLD in men (OR 0.87) and increased risk in women (OR 1.22) [27].
Therefore, although NAFLD seems to affect more males than females, age and menopausal status play an important role in sex inequalities in NAFLD prevalence. Moreover, even though several studies refer to this SDH as gender, the analyses are based on biological sex (male vs. female) only.

Ethnicity/Race
The exact prevalence of ethnicity differs according to the methodology employed and the population investigated. Overall, the prevalence of NAFLD appears to be higher among Hispanics, followed by non-Hispanic Whites and Asians, and lastly, African Americans in the US population [26,[29][30][31]. Furthermore, this ethnic disparity has also been observed in hospitalized patients [32] and lean-NAFLD patients [23].
In addition, there are substantial differences in the mean age of NAFLD onset by race. Although it seems that Mexican Americans have an earlier NAFLD onset (36.7 years) [30], it is not clear whether African Americans or non-Hispanic Whites have the latest onset. Schneider et al. observed that the mean age for Non-Hispanic Whites is 43.1 years, while the mean age for African Americans is 39.5 years [30]. Another study suggested that African Americans with NAFLD were significantly older than other racial or ethnic groups [31].
Although Hispanic ethnicity has the highest NAFLD prevalence, Japanese Americans have a greater susceptibility to intra-abdominal adiposity, which is a risk factor for NAFLD [28]. Even though Hispanics have a higher visceral fat mass, once the total mass is accounted for, their visceral fat mass proportion is similar to that of Whites [28]. There are also differences in liver enzymes such as aspartate and alanine transaminases levels, which are used as NAFLD biomarkers. Although AST and ALT levels are the highest among Hispanics, African Americans have significantly higher mean AST levels compared to non-Hispanic Whites [30]. Hence, although the adiposity body distribution and AST/ALT levels differ between ethnicities, they do not follow the same prevalence pattern as NAFLD.
Hispanics appeared to have a higher NAFLD prevalence, a sooner onset, and a worse metabolic profile than other ethnicities. Of note, all the studies that analyze ethnic disparities were performed among the US population.

Socioeconomic Position/Status
Overall, NAFLD prevalence seems to be higher among individuals with lower socioeconomic position in Western countries [33,34]. However, data coming from Eastern countries is not so consistent [35,36]. A study conducted in South Korea found that people with a low socioeconomic status have a significantly higher risk of developing NAFLD (OR 1.7) [36]. A Chinese study also found that people with a higher median income had a 1.96 higher risk of developing NAFLD than the low-income population [35]. However, this study had a notable selection bias that limits external validity [35].
In the field of NAFLD, the socioeconomic position has been particularly studied by addressing how it relates to food insecurity and the limited or uncertain access to nutritionally adequate and safe foods [33,34]. Approximately 29% of US adults in lowincome households with NAFLD live in food-insecure households [33]. Subjects who live under food insecurity have higher odds of developing NAFLD (OR 1.38) [33]. Similarly, in the Iranian population, the prevalence of food insecurity is much higher among the NAFLD population (56.8%) than among individuals without NAFLD (26.1%) [34].

Advanced Liver Disease in NAFLD Gender/Sex
Although a global population study concluded that among the population below 70 years, males had a higher prevalence of cirrhosis and several DALYs (Disability-Adjusted Life Years) because of liver disease [38]. In the US, women seem to have a higher prevalence of NASH-related cirrhosis than males, both compensated and decompensated [32,39]. Little has been studied about gender/sex differences in NAFLD progression to advanced fibrosis; however, it seems that the prevalence of advanced fibrosis is similar (OR 1.6 vs. 1.4) when only postmenopausal women are analyzed [40]. Furthermore, among females, the elderly have higher rates of advanced fibrosis at NAFLD onset (OR 1.8) [40].

Ethnicity/Race
No studies analyzing the impact of ethnicity/race as a risk factor of either advanced liver fibrosis or cirrhosis nor the comparative risks or burdens in NAFLD patients were found.

Socioeconomic Position/Status
Advanced liver disease is more prevalent in individuals with lower socioeconomic status [33,36,38]. In geographic regions with an overall lower socioeconomic position, there is higher cirrhosis prevalence [38]. Furthermore, among low-income households, food insecurity is associated with higher odds of advanced fibrosis (OR 2.20) [33]. Elderly females (≥65 years) are more likely than elderly males to have NAFLD/NASH-related HCC (28.0% vs. 14.8%, p = 0.0006).
Older females with HCC have more NAFLD/NASH and may be overlooked by current surveillance guidelines.
Includes other liver diseases. Does not perform an intersectional analysis. Food insecurity may be independently associated with NAFLD and advanced fibrosis among low-income adults in the US.
Does not perform an intersectional analysis. Does not perform an intersectional analysis. The frequency of underlying NAFLD was significantly higher in women than men.
They do not classify women as pre/PMP. Includes other liver diseases. Hispanics have similar or better long-term post-LT outcomes compared to non-Hispanic whites despite a worse pretransplant risk factor profile.
Only analyzes Hispanics and non-Hispanic Whites.

NAFLD-Related Complications Gender/Sex
Regarding liver cancer, in a Chinese cohort of 1206 patients with HCC, Wu et al. found that just 25% were women, who were more likely to undergo HCC surveillance, had smaller tumor size at diagnosis, and less vascular involvement [30]. However, 21.5% of females had NAFLD as underlying liver disease, whereas only 7.2% of males did. Moreover, males who met the Milan criteria for liver transplantation were more likely to undergo liver transplant than women who met the criteria [30]. In an American cohort including 5327 patients with HCC, 22.6% were women, who had a significantly higher frequency of NAFLD (23% vs. 12%) and lower frequency of alcoholic liver disease (5% vs. 15%). The proportion of noncirrhotic HCC was higher among women (17% vs. 10%). Women had less-advanced HCC at presentation by tumor, node, metastasis staging, a higher proportion within Milan criteria (39% vs. 35%), and a greater overall survival (2.5 ± 2.9 years vs. 2.2 ± 2.7 years) [41].
NAFLD-related mortality according to sex differs across studies. Adejumo and colleagues found that in the US from 2007 to 2014 women had a significantly lower in-hospital mortality [32]. However, this study was based on ICD-9 diagnosis code, whose ability to capture NAFLD-related diagnosis and complications is severely limited. Meanwhile, in a population of 76,149 patients listed for liver transplant between 2005 and 2012 of whom 7.2% were listed for NASH, Loy et al. found that women with NASH cirrhosis had a higher risk of death on the liver transplant waiting list [42].
NASH is the main cause of liver transplant among females and the second one in males, right after alcoholic liver disease [6,42], yet the US women with NASH are 19% less likely to receive a liver transplant compared to men [42].
Ethnicity/Race NASH has increased as a cause of liver transplant in all ethnic subgroups. Ethnic inequalities have been observed in NAFLD-related complications such as liver transplant and hospitalization events. In 2016, NASH was the leading liver transplant cause among Asian, Hispanic, and non-Hispanic White females [6]. According to the lowest NASH prevalence, African American population had a lower increase in the rate of waitlist registration for NASH compared to the other ethnic groups [6]. Additionally, although the highest NASH prevalence has been seen among the Hispanic population [23,26], non-Hispanic White patients are more commonly transplanted for NASH compared to Hispanics in the US [43]. Although seemingly the lowest NASH prevalence is found among the African American population [23,26], these have a 14% greater odds of unfavorable discharge compared with whites [32].

Socioeconomic Position/Status
Patients with low socioeconomic status have higher liver cancer rates [31]. The odds of dying on the waitlist increase by 14% in NASH patients without health insurance or self-paid [32].

Discussion
NAFLD distribution and outcomes show large inequalities by social group. Males have higher NAFLD prevalence, while available evidence suggests that women have an overall higher risk of progression to advanced liver disease and more complications, whereas they are less likely to undergo transplantation once cirrhosis develops. In the US, the Hispanic population has the highest NAFLD prevalence and poorer outcomes, whereas Whites have fewer complications than other ethnicities. Patients with low socioeconomic status have higher NAFLD prevalence and higher likelihood of progression and complications. Globally, there is a lack of studies analyzing NAFLD prevalence and outcomes according to various axes of inequality through joint intersectional appraisals, and most studies included in our review are based on the US population.
Although in the general population males have a higher prevalence, if only postmenopausal women are taken into account, the prevalence is similar between male and female individuals. In addition, females have a greater risk of progression to advanced liver disease. In a systematic review, Balakrishnan et al. found that although they have a lower risk of NAFLD than men, once NAFLD is established, women have a higher risk of advanced fibrosis, especially after age 50 [13]. Although estrogens have been shown to play an important role as a protective factor against steatosis and therefore may at least partially account for an earlier NAFLD onset among men [15], little research has been carried out to elucidate such sex differences in NAFLD prevalence. Because females have lower healthcare access [44], they might be underdiagnosed or may receive a later diagnosis that can lead to further progression of the disease. Moreover, the fact that males tend to have a higher alcohol consumption than females [31] might predispose healthcare providers to rule out liver disease in men rather than women. Thus, aspects such as access to healthcare, living and working conditions, and gender roles should be included in further research exploring gender disparities in NAFLD.
Ethnicity/race is the axis of inequality that has been more largely documented (50% of analyzed studies addressed it), and it has been shown to have a greater influence on health inequalities in NAFLD prevalence and outcomes. In particular, the Hispanic population has a higher NAFLD prevalence and burden of advanced liver disease and complications. Yet, ethnic disparities in NAFLD distribution have mostly been studied among the US population [23,26,[29][30][31], and further studies should be conducted to determine if these inequalities are conserved in other countries. Ethnic disparities are not fully explained by the metabolic profile and the cardiometabolic comorbidities of the individuals, such as obesity, T2DM, or hypertension. For instance, although African Americans have the lowest prevalence, they have a worse cardiovascular profile [29] and similar diabetes and obesity prevalence as for other ethnicities [45,46]. Most studies suggest that ethnic disparities might be better explained by PNPLA3 polymorphisms, which are more frequent in the Hispanic population and promote fat accumulation on the liver [14]. Although genetic polymorphisms and metabolic profiles may play an important role in ethnic disparities, living and working conditions have a major impact on health inequalities beyond individual factors. The higher NAFLD prevalence among Hispanic ethnicity could be also explained by the fact that they have the worst health care access [47,48] and health insurance coverage [49]. Hence, further research is needed to understand the gene-environment interactions and epigenetic mechanisms including environmental and social factors determining NAFLD onset and progression.
Socioeconomic status is tightly associated with the risk of developing NAFLD. Our findings are consistent with previous studies that demonstrate that NCDs are more prevalent among the population with lower socioeconomic status [50]. Additionally, food insecurity seems to play an important role in NAFLD onset, as food-insecure households have been described as having a greater burden of cardiometabolic diseases, such as diabetes, obesity, and hypertension [51,52]. Food insecurity is characterized by difficult access to healthy food, which can lead to an early obesity onset in young ages that ends up with a metabolic syndrome. Therefore, since NAFLD is a metabolic disease, the increased NAFLD prevalence in the food-insecure population can be related to a higher prevalence of metabolic syndrome among this population [33]. In addition, low socioeconomic position is also associated with other individual outcomes such as depression and anxiety, or community conditions such as poor healthcare access or absence of safe spaces to exercise [51] that could also participate in inequalities in NAFLD distribution and outcomes.
NAFLD, like all non-communicable diseases, is a multifactorial disease that is influenced by both individual factors, such as behavior or genetics, and contextual factors such as culture, demographics, income inequality, environment, or health policies [51,53]. Nevertheless, most studies are only focused on the individual level, emphasizing selfresponsibility through a "victim-blaming" approach and ignoring the deeper social, economic, and environmental determinants of living and working conditions that can lead to NAFLD onset [54]. Even though genetic and metabolic factors have a paramount impact on NAFLD prevalence and progression, they do not explain all the observed inequalities. In the field of hepatology, social conditions are closely linked to liver disease prevalence and outcomes, as chronic liver disease disproportionately impacts racial and ethnic minorities, low socioeconomic position communities, and immigrants [55]. Therefore, NAFLD should be understood as context-sensitive and analyzed by taking into account the jointly influence of individual factors, such as genetics or behaviors, social factors such as culture, race, income, and education; environmental factors, such as availability and access to healthy foods, energy-dense foods or the proximity of safe walking spaces; and macro-level factors such as the influence of media and advertising, access to health care or governmental policies [51]. Hence, it is urgent to study NAFLD, as well as all NCDs, through approaches embedding multilevel analysis that allows examining simultaneously both communitylevel and individual-level SDH that can influence NAFLD onset and progression [56].
It should be noted that besides the scarcity of studies paying attention to SDH and inequalities in the field of NAFLD, there are two other relevant barriers that hinder an accurate epidemiological approach. Although the definition of NAFLD relies on the presence of steatosis in at least 5% of hepatocytes, in reality, the approach to NAFLD diagnosis widely varies across studies, which largely impacts the reports of NAFLD prevalence. Moreover, whereas the concept of MAFLD has been proposed by some experts to facilitate the clinical approach to fatty liver disease in patients with relevant metabolic disorders [57], it is a potential source of heterogeneity both in clinical and epidemiological studies. By neither excluding patients with alcohol consumption over the dose traditionally considered for risk for liver injury nor those with concomitant liver disease, MAFLD widely differs from NAFLD. Hence, data from studies reporting on MAFLD distribution and outcomes should be carefully analyzed, and direct extrapolation between MAFLD and NAFLD should be avoided.
This study has several limitations. First, this is the first review on a vast topic and it only explored the PubMed database, and only articles written in English were included, so some articles may have been missed. Yet, scoping reviews are a rigorous, transparent, and replicable method that provide a robust assessment of the potential size and scope of available research literature and allow for identifying knowledge gaps when there is not enough evidence in the field to raise a specific question that can be studied through a systematic review [16]. Second, most studies addressing advanced liver disease in NAFLD also include other causes of liver disease, such as viral hepatitis or alcoholic liver disease. Thus, it is difficult to extract data that include only NAFLD/NASH as a cause of liver disease. Third, we have not analyzed some axes of inequalities that can have an important influence on health disparities, such as migration status or other axes. Fourth, most of the included studies do not specify how gender has been described and may be referring to biological sex as gender. However, we acknowledge that part of the differences between men and women on NAFLD are related to gendered social roles and not to biological features. Additionally, although gender is a spectrum of many genders, we have only analyzed male and female identities, as they are the most studied ones. Fifth, most studies included in our review concern the US population, therefore limiting the external validity; e.g., some geographical settings such as South America and some Asian countries appear to have higher burdens of NAFLD [4], but data are lacking on its distribution according to SDH. Fifth, we decided to include only original articles and therefore excluded both metaanalyses and reviews from our study, which might have led to the omission of relevant studies addressing at least one of the axes of inequality in the distribution and outcomes of NAFLD [58]. Finally, the included studies do not specifically analyze the impact of social determinants of health on the onset of health inequalities, and most of them do not assess the intersection of more than one axis of inequality. Hence, more research should be conducted to understand how the interaction between different social determinants of health (e.g., employment and neighborhood environment) leads to social health inequalities in NAFLD.

Conclusions
Hispanics, males, and people with low socioeconomic position present higher rates of NAFLD, whereas females present relatively poorer outcomes. Overall, little has been studied on the relationship between SDH and NAFLD. Moreover, most research has been carried out in the US; hence, evidence to provide an accurate assessment of the relationship between SDH and NAFLD in many regions is lacking. In addition, intersectional approaches encompassing several SDH and axes of inequality that provide an integrated account of both the biomedical and social determinants of NAFLD are needed. Further research is warranted to shed light on the impact of SDH on NAFLD distribution and outcomes both at the national and global levels.