Shifting perspectives – interplay between non-alcoholic fatty liver disease and insulin resistance in lean individuals

Table 1 Prevalence, characteristics, and outcomes in lean/non-obese individuals with non-alcoholic fatty liver disease

Author, year	Population, study design, sample size	Prevalence of NAFLD in lean subjects	Main findings
Zou et al[4], 2020	Mixed population, 1999-2016 NHANES databases	32.3% overall NAFLD prevalence; 22.7% obese and 9.6% non-obese; Amongst NAFLD patients, 29.7% were non-obese (Caucasian BMI 25-30 kg/m2, Asian BMI 23-27 kg/m2), of which 13.6% had lean NAFLD (Caucasian BMI < 25 kg/m2, Asian BMI < 23 kg/m2)	Non-obese NAFLD individuals had higher 15-year cumulative all-cause mortality (51.7%) than obese NAFLD (27.2%) and non-NAFLD (20.7%)
Huang et al[20], 2020	2483 Asian participants, community based study	44.5% NAFLD and 15.8%, MetS prevalence; Among NAFLD subjects, 48.8%were obese (BMI ≥ 24 kg/m2)	IR is predictive of NAFLD irrespective of BMI; CV risk calculated by Framingham Risk Score may exist in lean NAFLD subjects
Tobari et al[18], 2020	Asian, biopsy-proven 762 NAFLD patients, cross sectional study	Over 25% men and almost 40% women were non-obese, but most of them had visceral fat obesity and/or IR; BMI cutoff 25 kg/m2	NAFLD was not milder in non-obese patients; Histological steatosis was associated with BMI; Advanced fibrosis was not associated with BMI and showed a significant sex difference
Kim et al[10], 2020	664 Asian subjects with biopsy-proven NAFLD and controls, cross sectional study	542 subjects with biopsy-proven NAFLD132 non-obese NAFLD (BMI < 25 kg/m2) ; 410 obese NAFLD (BMI > 25 kg/m2) ; 122 controls	Non-obese subjects with NAFLD displayed a similar severity of histological liver damage; Sagittal abdominal diameter was independently associated with significant fibrosis among subjects with non-obese NAFLD
Alferink et al[71], 2019	4609 elderly European, population based study	1623 had NAFLD (n = 161 normal-weight and n = 1462 overweight, BMI cutoff 25 kg/m2)	Both high fat mass and low SMI were associated with normal-weight NAFLD; Fat distribution (assessed by AGR) could best predict NAFLD prevalence
Denkmayr et al[19], 2018	European, 466 patients diagnosed with NAFLD, cross sectional study	Lean (BMI ≤ 25.0 kg/m², n = 74) ; Overweight (BMI > 25.0 ≤ 30.0 kg/m², n = 242) ; Obese (BMI > 30.0 kg/m², n = 150)	Lean NAFLD patients had a histological picture similar to obese patients but more severe compared to overweight patients.
Gonzalez-Cantero et al[21], 2018	European, cross-sectional study 113 non-obese, non-diabetic individuals	55 patients diagnosed with NAFLD; NAFLD defined as hepatic triglyceride content > 5.56% (quantified by 3T H1-MRS) ; BMI cutoff 25 kg/m2	Lean-with-NAFLD group had significantly higher HOMA-IR and lower serum adiponectin than the overweight-without-NAFLD group; IR was independently associated with NAFLD but not with waist circumference or BMI
Hagström et al[5], 2017	European, prospective cohort study of 646 patients with biopsy-proven NAFLD	19% lean NAFLD; 52% overweight NAFLD; 29% obese NAFLD; BMI cutoff 25 and 30 kg/m2	Lean NAFLD had lower stages of fibrosis and higher risk for severe liver disease development compared to patients with NAFLD and a higher BMI, independent of available confounders (follow-up 19.9 years)
Leung et al[6], 2017	Asian, prospective, 307 NAFLD patients	23.5% were non-obese; BMI cutoff 25 kg/m2	Non-obese NAFLD patients have less-severe disease and may have a better prognosis than obese patients; Hypertriglyceridemia and higher creatinine are the key factors associated with advanced liver disease in non-obese patients
Fracanzani et al[11], 2017	European, retrospective cohort study of 669 patients with biopsy-proven NAFLD	143 patients had BMI < 25 kg/m2 and NAFLD	20% of patients with lean NAFLD have NASH, fibrosis scores of 2 or higher, and carotid atherosclerosis
Feldman et al[22], 2017	Caucasian, cross sectional, 187 subjects with hepatic steatosis on ultrasound	Lean healthy (BMI ≤ 25 kg/m2, no steatosis, n = 71) ; Lean NAFLD (BMI ≤ 25 kg/m2, steatosis, n = 55) ; obese NAFLD (BMI ≥ 30 kg/m2, steatosis; n = 61)	Lean NAFLD have impaired glucose tolerance, low adiponectin concentrations and an increased rate of PNPLA3 risk allele carriage
Feng et al[7], 2014	Asian, population based, 1779 participants	The prevalence of NAFLD was 18.33% in the lean group and 72.90% in the overweight-obese groupBMI cutoff 24 kg/m2	Lean-NAFLD was more strongly associated with diabetes, hypertension, and MetS than overweight-obese-NAFLD; NAFLD patients were more likely to have central obesity especially in lean groups
Younossi et al[17], 2012	Mixed population, 1988-1994 NHANES databases	2185 (18.77% ± 0.76%) of subjects had NAFLD; 7.39% ± 0.65% had lean NAFLD; 27.75% ± 1.00% had overweight/obese NAFLDBMI cutoff 25 kg/m2	Lean NAFLD was independently associated with younger age, female sex, and a decreased likelihood of having IR and hypercholesterolemia
Margariti et al[16], 2012	European, cross sectional, 162 NAFLD patients	Normal BMI was present in 12% of patients; BMI cutoff 25 kg/m2	Lean NAFLD patients do not have IR-associated metabolic disorders, but they have higher levels of ALT/AST than the overweight or obese NAFLD patients

3T H1-MRS: 3Tesla H1-magnetic resonance spectroscopy; ALT: Alanine aminotransferase; AGR: Android gynoid ratio; AST: Aspartate aminotransferase; BMI: Body mass index; CV: Cardiovascular; IR: Insulin resistance; MetS: Metabolic syndrome; NAFLD: Non-alcoholic fatty liver disease; NASH: Non-alcoholic steatohepatitis; SMI: Skeletal muscle index.

Table 2 Definitions of metabolic health in non-obese

Definitions of metabolic health in non-obese individuals:
Absence of insulin resistance	Meigs et al[62]; Stefan et al[63]
Absence of insulin resistance and low CRP levels as a surrogate marker for inflammation, in combination with up to any two parameters of metabolic syndrome	Wildman et al[64]; Karelis et al[65]
Combination with up to any two parameters of metabolic syndrome	Stefan et al[66]; Phillips[67]
Definition of metabolically unhealthy non-obese individuals:
BMI < 25 kg/m2 and presence of insulin resistance	Stefan et al[23]
Waist circumference adjusted for BMI and/or android gynoid ratio and presence of insulin resistance	Suggested by authors

BMI: Body mass index; CRP: C-reactive protein.