Nonobese nonalcoholic fatty liver disease

Watch a video presentation of this article

Watch the interview with the author

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BMI

body mass index

NAFLD

nonalcoholic fatty liver disease

NASH

nonalcoholic steatohepatitis

PNPLA3

palatin‐like phospholipase domain‐containing 3

SREBF‐2

sterol regulatory element‐binding factor 2

Nonalcoholic fatty liver disease (NAFLD) is now the most common cause of liver disease in the Western world, and its prevalence is growing worldwide.1 NAFLD can progress from simple steatosis without inflammation or fibrosis to nonalcoholic steatohepatitis (NASH) and cirrhosis. Typically, NAFLD is associated with metabolic syndrome, including obesity, lipid abnormalities, and insulin resistance. The prevalence of NAFLD increases by a factor of 4.6 in obese people, defined by body mass index (BMI) >30 kg/m².2 The prevalence of obesity in patients with NAFLD varies between 30% and 100%.3 However, there is growing evidence to show the significant prevalence of NAFLD in nonobese patients. This review focuses on the prevalence, clinical characteristics, and histology of nonobese NAFLD in adult patients.

Obesity: Definition and Measurement

Obesity is difficult to define. Waist circumference and BMI are often used to define obesity; however, both are flawed markers of adiposity. BMI does not account for body fat distribution, which is thought to be more important than body fat content.4 Waist circumference is used as a surrogate marker for visceral adiposity, but waist measurements correlate more with subcutaneous fat than with visceral fat.5 Visceral fat is more relevant than subcutaneous fat because it plays a greater role in the pathogenesis of insulin resistance.6

There are racial and ethnic differences in the associations between BMI, body fat percentage, and metabolic health, which has prompted cutoffs for overweight and obesity to differ between populations.7 There are suggested differences in distribution of central adiposity and visceral fat in Asian populations when compared with other ethnic groups, with greater central adiposity in Asian populations.7 Overweight is defined as ≥23 kg/m² for Asians and ≥25 kg/m² for other races, whereas obesity is defined as ≥25 kg/m² for Asians and ≥30 kg/m² for other races. Thus, nonobese NAFLD is defined as <25 kg/m² for Asians and <30 kg/m² for other races.

Although obesity and metabolic syndrome are often seen together, a subset of individuals are obese but do not have metabolic complications; they are defined as the “metabolically healthy obese.” There is also a group of individuals with insulin resistance, as well as other components of metabolic syndrome, but with a normal BMI; these individuals are described as metabolically obese normal weight.

II. Prevalence of Nonobese NAFLD

The reported global prevalence rate of nonobese NAFLD varies widely, ranging from 3% to almost 30%.1 The prevalence rate of obese NAFLD ranges from 28% to 60%.8, 9, 10, 11 The variation in prevalence between studies is due to differences in the definition of obesity across populations and the different modalities used to detect NAFLD, including imaging and histology.

III. Clinical Characteristics and Histology of Nonobese and Obese NAFLD

Several studies have investigated the epidemiology of non‐NAFLD, and several factors are associated with nonobese NAFLD (Table 1). Wei et al.8 conducted a population study investigating the prevalence and severity of NAFLD among nonobese Chinese people, defined as BMI <25 kg/m². They evaluated subjects with transient elastography and proton‐magnetic resonance spectroscopy, which correlates with hepatic steatosis on histology. Obese patients with NAFLD are more likely to have hypertension, insulin resistance, metabolic syndrome, higher degree of liver stiffness, and greater waist circumference when compared with nonobese patients with NAFLD.8 Nonobese patients with NAFLD, in contrast, are more likely to possess palatin‐like phospholipase domain‐containing 3 (PNPLA3) G polymorphisms.8

Table 1.

Clinical Characteristics That Differ Between Obese Patients With NAFLD Compared With Nonobese Patients With NAFLD1, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18

Characteristics	Obese NAFLD	Nonobese NAFLD
Age	↑
AST and ALT levels	↑
Blood pressure	↑
BMI	↑
Cytokeratin 18 fragment levels	↑
Fibrosis stage	↑
Insulin resistance	↑
Metabolic syndrome	↑
Necroinflammatory activity grade	↑
Liver stiffness	↑
Proportion with PNPLA3 polymorphisms		↑
Total cholesterol	↑
Triglycerides	↑
Waist circumference	↑

Several independent factors are associated with obese NAFLD when compared with nonobese NAFLD. Younossi et al.9 investigated factors that are associated with nonobese NAFLD in a population of patients in the United States. Multivariate analysis showed that nonobese NAFLD, as defined using ultrasound, was independently associated with female sex, young age, decreased insulin resistance, and decreased likelihood of hyperlipidemia when compared to overweight‐obese NAFLD patients.9 These results are similar to work from Belgium, which showed that nonobese patients with NAFLD are younger, and have less hypertension when compared with those with obese NAFLD.12 However, when compared with healthy controls, patients with nonobese NAFLD had more insulin resistance and more hypertriglyceridemia12; these results are similar to other studies (Table 2).13, 14, 15 There is some thought that plasma triglycerides are a proxy for lipotoxicity, which occurs when insulin‐resistant adipose tissue releases fatty acids, leading to increased lipoapoptosis and inflammation and hepatic steatosis.12

Table 2.

Clinical Characteristics That Differ Between Nonobese Patients With NAFLD and Nonobese Patients Without NAFLD1, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18

Characteristics	Nonobese With NAFLD	Nonobese Without NAFLD
Age	↑
AST and ALT	↑
Blood pressure	↑
BMI	↑
Hispanic ethnicity	↑
Insulin resistance	↑
Proportion with PNPLA3 polymorphisms	↑
Total cholesterol	↑
Triglycerides	↑
Waist circumference	↑

The clinical characteristics independently associated with obese NAFLD when compared with nonobese NAFLD are similar between Asian and non‐Asian countries. Das et al.10 examined the risk factors for nonobese NAFLD in a community‐based study in India and found that patients with nonobese NAFLD have a higher BMI, bicep skinfold thickness, and visceral adiposity compared with those without NAFLD but similar body weight. Further work by Kim et al.11 examined more than 700 patients older than 30 years of age in Korea and suggested that male gender, waist circumference, hypertriglyceridemia, and insulin resistance, as calculated by the homeostasis model assessment method, were independently associated with NAFLD in nonobese patients. A recent systematic review and meta‐analysis of 16 studies showed that patients with lean NAFLD, defined as BMI ≤25 kg/m², had lower fasting glucose, insulin resistance, systolic and diastolic blood pressure, BMI, waist circumference, and laboratory values (including aspartate aminotransferase [AST], alanine aminotransferase [ALT], and Gamma‐Glutamyl Transferase [GGT]) compared with patients with obese NAFLD (BMI >25 kg/m²).17

In terms of histological differences between nonobese and obese NAFLD, work by Leung et al.18 showed that nonobese patients with NAFLD had lower NAFLD activity scores, lower fibrosis stage, and lower liver stiffness measurement by transient elastography compared with obese patients with NAFLD; however, there were similar degrees of lobular inflammation between groups. Once NASH developed, however, there was no significant difference in advanced fibrosis (F3‐F4).18 Previous work has also shown that the prevalence of NASH and the severity of fibrosis between nonobese and obese NAFLD is similar.13, 14, 15

Genetics

Many risk factors are implicated in the development of nonobese NAFLD, including visceral adiposity and insulin resistance, but there are also genetic factors that may play an important role in nonobese NAFLD. The PNPLA3 polymorphism has been shown to be an independent risk factor for nonobese NAFLD after adjusting for other covariates, including insulin resistance.19 Specifically, the G allele in PNPLA3 increases the risk for NAFLD in patients with metabolic syndrome independent of metabolic factors, and heterozygote patients are at intermediate risk.20 In addition, there are some data to suggest that sterol regulatory element‐binding factor 2 (SREBF‐2) is associated with nonobese NAFLD in patients without metabolic syndrome or diabetes.21 Patients with SREBF‐2 also have a higher likelihood of NASH.21 The cholesteryl ester transfer protein, a plasma lipid transport protein, has also been associated with increased risk for fatty liver in nonobese patients.22 Further work is needed to understand the genetics risk factors for nonobese NAFLD and how the identification of these risk factors can be used to change clinical outcomes.

Conclusion

There is increasing interest in the prevalence and consequences of NAFLD; however, little attention has been given to NAFLD in nonobese individuals given it is typically thought of in association with obesity and metabolic syndrome. The prevalence of nonobese NAFLD is estimated to be up to 30% worldwide, and there is risk for progression from NAFLD to NASH and cirrhosis. It is important for clinicians to understand what characteristics put nonobese patients at risk for NAFLD given that nonobese NAFLD is not rare or benign.