TYPE Original Research PUBLISHED 13 January 2023 DOI 10.3389/fendo.2022.1051958 Abdominal obesity and OPEN ACCESS dsyglycemia are risk factors EDITED BY Andreea Ciudin, Vall d’Hebron University Hospital, for liver fibrosis progression Spain REVIEWED BY in NAFLD subjects: A Jesús Rivera Esteban, Vall d’Hebron University Hospital, Spain population-based study Diego Bellido Guerrero, Servicio Gallego de Salud, Spain Marı́a Teresa Julián 1†, Sı́lvia Ballesta 1†, Guillem Pera 2, *CORRESPONDENCE Núria Alonso Alejandra Pérez-Montes de Oca 1, Berta Soldevila 1,3, nurialonsopedrol@gmail.com Llorenç Caballerı́a 2,4, Rosa Morillas 4,5, Carmen Expósito 2,4, Dı´dac Mauricio didacmauricio@gmail.com Alba Martı́nez–Escudé 2,6, Manel Puig-Domingo 1,3,7, † These authors have contributed Josep Franch-Nadal 7,8, Pere Torán 2,4, Kenneth Cusi 9, equally to this work and share Josep Julve 7,10,11, Dı́dac Mauricio 7,10,12*‡ and Núria Alonso 1,3,7*‡ first authorship ‡ These authors share last authorship 1 Department of Endocrinology and Nutrition, Hospital Universitario Germans Trias I Pujol, Badalona, Barcelona, Spain, 2 Unitat de Suport a la Recerca (USR) Metropolitana Nord, Fundació SPECIALTY SECTION Institut Universitari d’Investigació en Atenció Primària Jordi Gol i Gurina (IDIAP Jordi Gol), Mataró, This article was submitted to Spain, 3 Department of Medicine, Universitat Autònoma de Barcelona, Barcelona, Spain, 4 Centro Translational Endocrinology, d’Investigaciones Biomédicas en Red, Enfermedades Hepatologia y Digestivas, Barcelona, Spain, a section of the journal 5 Hepatology Department, Hospital Germans Trias I Pujol, Badalona, Barcelona, Spain, 6 Primary Frontiers in Endocrinology Care, Centre d’Atenció Primària La Llagosta, Institut Català de la Salut, Barcelona, Spain, 7 Center for Biomedical Research on Diabetes and Associated Metabolic diseases (CIBERDEM), Instituto de RECEIVED 23 September 2022 Salud Carlos III, Madrid, Spain, 8 Primary Health Care Center Raval Sud, Gerència d’Atenció Primaria, ACCEPTED 21 December 2022 Institut Català de la Salut, Barcelona, Spain, 9 Division of Endocrinology, Diabetes and Metabolism. PUBLISHED 13 January 2023 University of Florida, Gainesville, FL, United States, 10Department of Endocrinology and Nutrition, CITATION Hospital de la Santa Creu i Sant Pau; Institut d'Investigació Biomédica Sant Pau (IIB Sant Pau), Julián MT, Ballesta S, Pera G, Barcelona, Spain, 11 Department of Biochemistry and Molecular Biology, Universitat Autònoma de Pérez-Montes de Oca A, Soldevila B, Barcelona, Barcelona, Spain, 12 Faculty of Medicine, University of Vic, Central University of Catalonia, Caballerı´a L, Morillas R, Expósito C, Vic, Spain Martı´nez–Escudé A, Puig-Domingo M, Franch-Nadal J, Torán P, Cusi K, Julve J, Mauricio D and Alonso N (2023) Abdominal obesity and dsyglycemia are risk factors for liver Objective: To investigate longitudinal changes in the liver stiffness fibrosis progression in NAFLD measurement (LSM) in the general adult population without known liver subjects: A population-based study. disease and to describe its association with metabolic risk factors, with a Front. Endocrinol. 13:1051958. doi: 10.3389/fendo.2022.1051958 special focus on subjects with non-alcoholic fatty liver disease (NAFLD) and dysglycemia. Material and Methods: A longitudinal adult population-based cohort study was conducted in Catalonia. LSM was measured by transient elastography (TE) at baseline and follow-up (median: 4.2 years). Subgroup with NAFLD and dysglycemia were analyzed. Moderate-to-advanced liver fibrosis was defined as LSM ≥8.0 kPa and LSM ≥9.2 kPa respectively. Results: Among 1.478 subjects evaluated, the cumulative incidence of LSM ≥8.0 kPa and ≥9.2 kPa at follow-up was 2.8% and 1.9%, respectively. This incidence was higher in NAFLD (7.1% for LSM ≥8.0 kPa and 5% for LSM ≥9.2 kPa) and dysglycemia (6.2% for LSM ≥8.0 kPa and 4.7% for LSM ≥9.2 kPa) subgroups. In the global cohort, the multivariate analyses showed that dysglycemia, Frontiers in Endocrinology 01 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 abdominal obesity and atherogenic dyslipidemia were significantly associated with progression to moderate-to-advanced liver fibrosis. Female sex was negatively associated. In subjects with NAFLD, abdominal obesity and dysglycemia were associated with changes in LSM to ≥8.0 kPa and ≥9.2 kPa at follow-up. A decline in LSM value to <8 kPa was observed in 64% of those subjects with a baseline LSM ≥8.0 kPa. Conclusions: In this population study, the presence of abdominal obesity and dysglycemia were the main risk metabolic factors associated with moderate- to-advanced liver fibrosis development over time in general populations as well as in subjects with NAFLD. KEYWORDS liver fibrosis, NAFLD, Transient elastography, abdominal obesity, dysglycemia, general population Introduction disease. In addition, we also aimed to take the same approach in subjects with NAFLD and dysglycemia. Progressive liver fibrosis is the main determinant of long- term outcome in chronic liver diseases (CLDs). Although liver biopsy is considered the gold standard for assessing hepatic Material and methods fibrosis, liver stiffness measurement (LSM) by transient elastography (TE) is one of the most widely used tools (1). Study design and participants LSM has been used to measure fibrosis in patients with various types of CLDs, including non-alcoholic fatty liver disease This was a longitudinal population-based study based on a (NAFLD). However, few studies have investigated TE as a previous cross-sectional descriptive population study (5, 6). The screening tool in the general population to identify study was conducted in several municipalities in the northern asymptomatic subjects with liver fibrosis. Data from part of the Barcelona metropolitan area. The recruitment period population-based studies revealed that 4-7% of subjects was between April 2012 and January 2016, while a second cross- without known liver disease have significant liver fibrosis sectional evaluation was performed between October 2016 and assessed by TE, mostly associated with NAFLD (2–6). In December 2019. Participants in the study were randomly addition, several studies have evaluated the risk factors identified from a total of 162,950 subjects aged 18–75 years associated with significant liver fibrosis (2, 3). In a previous from the registries of the primary health care centers of the cross-sectional population-based study, we described that factors municipalities included in the study. Randomly-identified independently associated with increased LSM were male gender, subjects were contacted through telephone and invited to abdominal obesity, type 2 diabetes (T2D) and dyslipidemia (6). participate. Patients with a current history of liver disease, To date, we are not aware of studies that have evaluated including cholestasis, hepatitis C or B virus infection, and longitudinal changes in LSM assessed by TE in the general high-risk alcohol consumption were excluded from the study. population. Given that the natural history of CLD depends Other exclusion criteria were active malignancy, other severe mainly on the underlying cause (7), most studies that evaluate diseases (congestive heart failure New York Heart Association the rate of fibrosis progression do it in a group of patients with >2, chronic obstructive pulmonary disease defined as Global the same CLD, with or without a specific treatment (7–9). The Initiative for Chronic Obstructive Lung Disease score >2, liver stiffness value by TE, as a non-invasive tool, may be useful chronic kidney disease requiring dialysis, previous organ for identifying progression to significant liver fibrosis over time transplantation, and severe neurological diseases) or admission in asymptomatic adult population. in long-term nursing homes. After obtaining informed consent, The aim of the present study was to investigate the the following steps were taken: (i) detailed medical history, longitudinal changes of LSM assessed by TE and its associated including alcohol consumption, anthropometric metabolic risk factors within a well-characterized large measurements, including body weight, height, body mass Mediterranean population-based cohort without known liver index (BMI), waist circumference, and arterial pressure. Frontiers in Endocrinology 02 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 Abdominal obesity was described as waist circumference ≥102 HDL-C in serum. Remnant cholesterol concentration cm in men or ≥88 cm in women; (ii) blood tests were performed (expressed as mg/dL) was calculated using the following after 12 hours of fasting, including liver biochemistry, hepatitis B equation: total cholesterol – LCL-C – HDL-C. and C virus markers (HBsAg and anti-HCV), serum fasting glycemia, glycosylated hemoglobin (HbA1c), serum creatinine, serum ferritin and serum lipid profile (total cholesterol (TC), Statistical analysis LDL cholesterol (LDL-C), HDL cholesterol (HDL-C), non-HDL cholesterol (non-HDL-C), cholesterol remnants, triglycerides Descriptive analysis used frequencies and percentages (TG) (iii) a TE with LSM was performed at baseline and at the (categorical variables), means and standard deviation end of the follow-up period. (symmetric distributed continuous variables), and median and interquartile range (skewed continuous variables). Comparison of baseline vs follow-up means was assessed by paired t-Student Definitions tests and comparison of percentages by McNemar tests. Univariate lineal regression was used to assess the The presence of NAFLD was defined using standard association between different metabolic risk factors and diagnostic criteria: one or more positive findings regarding progression of liver fibrosis (dependent variable). Progression fatty liver index (FLI ≥60), abdominal echography or liver was computed as the LSM at the end of follow-up minus LSM at biopsy. In our cohort, we excluded cases of viral hepatitis, high baseline. Univariate logistic regression was performed to obtain alcohol risk consumption and other liver disease. Thus, the the odds ratio (OR) and their 95% confidence intervals (95% CI) evaluation of changes in LSM was to be primarily affected by of these risk factors regarding the liver fibrosis progression, metabolic disorders. computed as changing, between the baseline and the end of High-risk alcohol consumption was defined by standard follow-up, the LSM category from less than 8 to 8 or more kPa drinking units/week; >21 in men and >14 in women. High (or, less than 9.2 to 9.2 or more kPa). Different multivariate serum transaminase concentrations were defined as aspartate regression models, both of lineal and logistic analysis, were transaminase (AST) and/or alanine transaminase (ALT) >40 IU/ obtained adjusting for potential confounders. Only variables L. The FLI, FIB-4 index and NAFLD fibrosis score (NFS) was with a p value less than 0.1 in the univariate models were calculated as previously published (10–12). The FIB-4 index and included in the adjusted models. The final multivariate models NFS were considered as high risk of significant fibrosis according were selected only containing variables with p <0.05 and, if to cut-off points described in the original publications, being several models had different significant variables, the model with FIB-4 >2.67 and NFS ≥0.676 respectively (11, 12). the best adjustment was selected. All comparisons were bilateral, TE was performed using the Fibroscan® system (402, and the significance was 0.05. Statistical analysis was performed Echosens®, Paris, France) by three trained specialist liver using the Stata v17 statistical package. nurses. The XL probe was not available, so only the M probe was used. Moderate-to-advanced liver fibrosis was defined by LSM ≥8.0 kPa according to others epidemiological studies. Ethics Moreover, LSM ≥9.2 kPa was adopted to describe advanced fibrosis (2, 3, 13). This cut-off has been shown to predict The protocol was concordant with current and relevant significant liver fibrosis (F2-F4) with a high sensitivy and guidelines and regulations and it was approved by the Ethics specificity in a previous large population-based study from Committee of the Fundació Gol i Gorina (P11/58) (Barcelona, Catalonian region (5) We defined progression or regression Spain). Written informed consent was obtained from from one stage to another (< 8 a ≥8 kPa and < 9.2 to ≥ 9.2 all participants. and viceversa) but requiring, at least, a change (increase or decrease) of 1 kPa between LSM measurement at baseline and follow-up, based on previous studies (9, 14). Results Type 2 diabetes mellitus and prediabetes diagnosis were based on a registered diagnosis in the clinical records or having Baseline characteristics of the an HbA1c ≥6.5% or fasting glucose ≥126 mg/dL and Hb1Ac 5.7- study population 6.4% or fasting glucose 100-125 mg/dL, respectively. Abdominal obesity was define as waist circumference ≥102 cm in men or From a total of 4,866 invited subjects, 3,076 accepted ≥88 cm in women. For analysis purposes, we included both T2D participation in the study (participation rate, 63.2%). After and prediabetes subjects under the predefined category of exclusion of subjects with previous chronic liver disease (n = dysglycemia. Atherogenic dyslipidemia (AD) was defined as 13), high-risk alcohol consumption (n = 155), hepatitis B or C having >150 mg/dL TG and <40/50 (men/women) mg/dL virus infection (n = 19) and not data available (n = 49), the final Frontiers in Endocrinology 03 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 number at baseline was 2,840. Of these, 1,363 (48%) were not Longitudinal changes in LSM and factors included in the second cross-sectional cut: 598 did not accept, associated within the whole study group 511 could not be located, 147 had moved from our area, and 107 were excluded for other reasons. Therefore, the final study Categorical changes in longitudinal LSM according to the stage of population consisted of 1,478 subjects with 2 LSM, conducted baseline TE (<8 kPa or ≥ 8 kPa and < 9.2 or ≥9.2 kPa) are represented at baseline and follow-up (median 4.2 years later; range, 3.0-5.5). in Figure 1. The cumulative incidence for liver fibrosis progression in Baseline and follow-up clinical and biochemical characteristics the whole follow-up was 2.8% for LSM ≥8.0 kPa and 1.9% for LSM of this cohort are shown in Table 1. The prevalence of moderate- ≥9.2 kPa. On the other hand, a decline in LSM to < 8 kPa was to-advanced liver fibrosis by TE (LSM ≥8.0 kPa and ≥9.2 kPa observed in 46 (64%) subjects with a baseline LSM ≥8.0 at follow-up. cut-off) was 4.7% (n = 70) and 2.5% (n = 37), respectively. The LSM value decreased from ≥9.2 kPa at baseline to <9.2 kPa over Comparison between subjects who completed and not the time in 23 (62%) subjects. Moreover, 19 (51%) subjects experienced a follow-up are represented in Supplementary Table 1. LSM decrease from ≥9.2 kPa to < 8 kPa (Supplementary Table 2A). TABLE 1 Characteristics of the study population at baseline and follow-up. Total (n = 1.478) NAFLD (n = 511) Dysglycemia (n = 365) Baseline Follow-up p Baseline Follow-up p Baseline Follow-up p Age, years 56 ± 11 60 ± 11 – 59 ± 9 63 ± 9 – 61 ± 8 65 ± 8 – Female, n (%) 918 (62) 918 (62) – 253 (50) 253 (50) – 190 (52) 190 (52) – Body mass index ≥ 30, kg/m2 28 ± 5 28 ± 5 <0.001 32 ± 4 32 ± 4 0.165 30 ± 5 30 ± 5 0.097 Abdominal obesity, n (%) 743 (51) 805 (55) <0.001 422 (83) 402 (79) 0.010 231 (64) 247 (69) 0.036 Dysglycemia, n (%) 365 (25) 393 (27) 0.044 214 (42) 230 (45) 0.095 365 (100) 282 (77) <0.001 Glucose, mg/dL 100 ± 24 100 ± 23 0.082 108 ± 31 110 ± 29 0.285 125 ± 35 122 ± 32 0.146 Glycated hemoglobin, % 5.7 ± 0.7 5.7 ± 0.7 0.004 5.9 ± 0.9 6.0 ± 0.9 0.042 6.4 ± 1.0 6.4 ± 1.0 0.522 Triglycerides, mg/dL 119 ± 71 115 ± 57 0.0121 161 ± 92 142 ± 65 <0.001 154 ± 96 140 ± 68 0.002 Total cholesterol, mg/dL 14 ± 38 208 ± 39 <0.001 216 ± 40 203 ± 44 <0.001 208 ± 39 197 ± 45 <0.001 LDL cholesterol, mg/dL 135 ± 33 130 ± 34 <0.001 135 ± 36 124 ± 38 <0.001 127 ± 35 118 ± 39 <0.001 HDL cholesterol, mg/dL 56 ± 13 55 ± 13 0.300 51 ± 11 50 ± 11 0.084 51 ± 12 51 ± 12 0.937 Cholesterol remnants†, mg/dL 23 ± 13 23 ± 14 0.517 30 ± 15 29 ± 18 0.071 29 ± 16 29 ± 21 0.924 Atherogenic dyslipidemia§, n (%) 123 (9) 137 (10) 0.237 90 (19) 89 (18) 0.910 66 (19) 71 (20) 0.522 ALT and/or AST >40 U/L, n (%) 105 (8) 102 (7) 0.798 69 (14) 56 (12) 0.128 43 (12) 29 (8) 0.035 FLI¶ 47 ± 28 49 ± 28 0.001 79 ± 13 74 ± 19 <0.001 65 ± 26 63 ± 26 0.147 FLI ≥60, n (%) 450 (35) 481 (37) 0.032 450 (96) 364 (78) <0.001 198 (59) 195 (58) 0.696 NAFLD ||, n (%) 484 (36) 677 (50) <0.001 484 (100) 415 (86) <0.001 205 (59) 245 (71) <0.001 Mean liver fibrosis by LSM (kPa) 4.8 ± 2.2 4.9 ± 2.0 0.048 5.8 ± 3.0 5.8 ± 2.8 0.842 5.8 ± 3.4 5.8 ± 3.1 0.961 Liver fibrosis by LSM ≥8.0 kPa, n (%) 70 (5) 63 (4) 0.448 62 (12) 54 (11) 0.346 43 (12) 38 (10) 0.456 Liver fibrosis by LSM ≥9.2 kPa, n (%) 37 (3) 41 (3) 0.572 34 (7) 37 (7) 0.655 26 (7) 28 (8) 0.715 FIB-4 >2.67, n (%) 22 (2) 51 (4) <0.001 11 (2) 24 (5) 0.005 6 (2) 15 (5) 0.007 High NFS, n (%) 15 (1) 73 (6) <0.001 11 (2) 55 (12) <0.001 14 (4) 48 (15) <0.001 Data are n (%) or mean ± SD. ALT, alanine aminotransferase; AST, aspartate aminotransferase; FLI, fatty liver index; GGT, gamma-glutamyltransferase; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LSM, liver stiffness measurement (by elastography); NAFLD, non-alcoholic fatty liver disease. † Cholesterol remnants: total cholesterol- LDL-C-HDL-C. ‡ Non-HDL cholesterol is defined as the difference between total and HDL-C. § Atherogenic dyslipidemia is defined by triglyceride >150 mg/dL and HDL-C <40 mg/dL in men and <50 mg/dL in women. ¶ FLI (fatty liver index) estimates the amount of fat in the liver and includes body mass index, waist circumference, and serum gamma-glutamyltransferase and triglycerides. || NAFLD (nonalcoholic fatty liver disease) was defined using standard diagnostic criteria (FLI, abdominal echography or liver biopsy. - Not included in the model. Frontiers in Endocrinology 04 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 A B FIGURE 1 Categorical changes in longitudinal LSM by stage at follow up. (A) Percentage of subjects who maintained stable, regressed or progressed LSM during the follow-up period for the LSM 8 kPa cut-off. The total cohort, the NAFLD subgroup and the dysglycemia subgroup are represented in bars. To consider a change, a minimum increase/decrease of 1 kPa was required. (B) Percentage of subjects who maintained stable, regressed or progressed LSM during the follow-up period for the LSM 9.2 kPa cut-off. The total cohort, the NAFLD subgroup and the dysglycemia subgroup are represented in bars. To consider a change, a minimum increase/decrease of 1 kPa was required. Frontiers in Endocrinology 05 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 Subsequently, the risk factors associated with progression in Longitudinal changes in LSM and mean LSM were analyzed. In the univariate lineal regression associated factors in NAFLD subjects analysis, factors associated with a higher increase in the mean LSM value at follow-up were dysglycemia, obesity and Among the entire cohort, 511 (36%) subjects met the criteria for abdominal obesity (Supplementary Table 3A). After NAFLD. Baseline and follow-up clinical and biochemical adjustment, the presence of abdominal obesity and characteristics are shown in Table 1. Changes in longitudinal dysglycemia were the two main conditions associated with LSM according to the stage of baseline TE (<8 kPa or ≥ 8 kPa significant increase in mean LSM during the follow-up period and < 9.2 or ≥9.2 kPa) are represented in Figure 1. The cumulative (Table 2A). Female sex was associated with a higher decline in incidence for liver fibrosis progression was 7.1% for LSM ≥8.0 kPa the mean LSM value over time. and 5.0% for LSM ≥9.2 kPa. At follow-up, a decline in LSM value Next, we investigated the factors positively associated with to < 8 kPa was observed in 39 (63%) subjects with baseline progression to moderate-to-advanced liver fibrosis. In the LSM ≥8.0 kPa. The LSM value decreased from ≥9.2 kPa at univariate logistic regression model, the metabolic factors baseline to <9.2 kPa in 62% (n=21/34) and to <8 kPa in 17 positively associated were: dysglycemia, obesity, abdominal (50%) subjects over time (Supplementary Table 2B). obesity, AD, cholesterol remnant and HDL-C (Supplementary In multivariate lineal regression, the presence of dysglycemia Table 4A). After multivariate analyses, dysglycemia, abdominal was associated with a significant increase in the mean LSM value obesity and AD, were significantly associated with LSM follow-up time (Table 2B). Factors independently associated progression for both cut-offs (≥8.0 kPa and ≥9.2 kPa), whereas with LSM progression to ≥8.0 kPa in the multivariate logistic female sex (ORs 0.3) was negatively associated (Table 3A). regression analysis were dysglycemia and abdominal obesity Mean changes in metabolic parameters (body weight, waist (Table 3B). When a cut-off of LSM ≥9.2 kPa was analyzed, circumference, plasma glucose concentrations and HbA1c) from dysglycemia and obesity were significantly associated with LSM baseline according to progression/regression in liver fibrosis at progression. Female sex was negatively associated with LSM follow-up are represented in Supplementary Table 5. progression to ≥8.0 kPa, but not with LSM ≥9.2 kPa. Interestingly, the mean weight gain among subjects that experimented progression to moderate-to-advanced liver fibrosis (LSM < 8.0 to ≥ 8 kPa or LSM <9.2 to ≥9.2 kPa), was Longitudinal changes in LSM and 3.2 and 3.5 Kg, respectively. On the other hand, an average associated factors in subjects weight loss of -2.5 and -5.3 kg was observed in those subjects with dysglycemia with decreased in LSM to < 8 kPa and < 9.2 kPa. Changes were also observed in the evolution of waist circumference, plasma Next, because we found that dysglycemia is one of the main glucose and HbA1c between those subjects who either showed a factors associated with liver fibrosis progression, we analyzed the progression/regression in LSM during follow-up. same endpoints in subjects with impaired glucose metabolism. TABLE 2 Multivariate analysis of risk factors associated with an increase in LSM over time. a) Total cohort 95% CI beta P value Dysglycemia 0.27 0.09 0.46 0.004 Abdominal obesity 0.27 0.11 0.43 0.001 Female sex -0.34 -0.50 -0.17 <0.001 b) NAFLD subgroup beta 95% CI P value Dysglycemia * 0.49 0.13 0.85 0.008 c) Dysglycemia subgroup beta 95% CI P value Abdominal obesity 0.46 0.02 0.90 0.039 CI, confidence interval; LSM, liver stiffness measurement (by elastography). Variables in each model mutually adjusted. *This is the only variable that is significant (p < 0.05), so it is a univariate model. Frontiers in Endocrinology 06 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 TABLE 3 Multivariate logistic regression analysis of risk factors associated with LSM ≥ 8 kPa and ≥ 9.2 kPa at follow-up. a) Total cohort LSM ≥ 8 kPa LSM ≥ 9.2 kPa OR 95% CI P value OR 95% CI P value Dysglycemia 2.4 1.2 4.7 0.013 3.1 1.4 7.3 0.008 Abdominal obesity 6.8 2.7 16.9 0.000 5.5 1.8 16.7 0.003 Female sex 0.3 0.1 0.5 0.000 0.3 0.2 0.8 0.012 Atherogenic dyslipidemia 2.4 1 5.4 0.040 2.8 1.1 6.9 0.026 b) NAFLD subgroup LSM ≥ 8 kPa LSM ≥ 9.2 kPa OR 95% CI P value OR 95% CI P value Dyglycemia 2.4 1.1 5 0.020 3.2 1.3 7.6 0.001 Abdominal obesity 5 1.1 22 0.034 – – – NS Female sex 0.5 0.2 0.9998 0,05 – – – NS Body mass index ≥ 30 – – – NS 5.3 1.2 22.7 0.026 c) Dysglycemia subgroup LSM ≥ 8 kPa LSM ≥ 9.2 kPa OR 95% CI P value OR 95% CI P value Abdominal obesity 10.5 2.3 47.4 0.002 9.8 1.3 75.3 0.028 Female sex 0.2 0.1 0.6 0.002 – – – NS CI, confidence interval; LSM: liver stiffness measurement (by elastography). NS: p > 0.05. Not included in the model. Variables in each model mutually adjusted. Among the entire cohort, 365 (25%) subjects had dysglycemia. large well-characterized population-based cohort, with special Baseline clinical and biochemical characteristics are shown in attention in analyzing those risks factors associated with Table 1. Changes in longitudinal LSM according to the stage of progression to moderate-to-advanced liver fibrosis. Of note, baseline TE (<8 kPa or ≥ 8 kPa and < 9.2 or ≥9.2 kPa) are we excluded secondary causes of liver disease, and thus the represented in Figure 1. The cumulative incidence was 6.2% for cause of liver fibrosis was primarily NAFLD. Several LSM ≥8.0 kPa and 4.7% for LSM ≥9.2 kPa. At follow-up, a decline community-based studies, including previous data from our in LSM value to <8 kPa was observed in 24 (56%) subjects with a current cohort (5, 6), have reported that the prevalence of liver baseline LSM ≥8.0 kPa (n = 31), and in 14 (54%) subjects with a fibrosis obtained by TE is high (ranging between 5.6% and 7.5%), baseline LSM ≥9.2 kPa (Supplementary Table 2C). mostly associated to NAFLD. This prevalence is higher among After adjustment, the presence of abdominal obesity was the individuals with associated metabolic risk factors, such as only factor associated with higher increase in the mean LSM diabetes (18%-27%) (6, 15–20). value (Table 2C) and with an increase in LSM to ≥8.0 kPa and In our cohort, using a cut-off value of LSM ≥8.0 kPa and ≥9.2 ≥9.2 kPa over time in a multivariate logistic regression analysis kPa by TE, the cumulative incidence of moderate-to-advanced (Table 3C). Female sex was negatively associated with an liver fibrosis was 2.8% and 1.9%, respectively, over a median increased in LSM but only for ≥8.0 kPa cut-off. period of around 4 years. Interestingly, extrapolation of our data to the European population between 18-75 years (312 million in 2018) indicates that approximately 8.7 million European citizens Discussion with unknown liver disease could develop significant liver fibrosis in a period around 4 years. In addition, as expected, To our knowledge, this is the first longitudinal study aimed when we focus on the subgroup of subjects with NAFLD or at investigating changes of LSM over time assessed by TE in a dysglycemia, the incidence of significant liver fibrosis at follow Frontiers in Endocrinology 07 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 up was higher than in the general population. In a previous Of note, in some liver conditions, such as alcoholic liver disease cross-sectional population study, we recently published that the or viral hepatic infection, effective treatment can change the prevalence of liver fibrosis assessed by TE was higher in subjects natural history of underlying CLD, resulting in a regression and/ with NAFLD (13% vs 1%) and T2D (21% vs 3.4%) compared or lower progression rate to liver fibrosis (23). Although no with those without these metabolic conditions (6). specific pharmacological treatment has yet been established for Further, we analyzed the metabolic risk factors associated fatty liver disease (23), it is known that a weight loss > 7% can with progression to moderate-to-advanced liver fibrosis during induce liver fibrosis regression (24). If we assume that in our follow-up period. In the current study, we found that abdominal cohort the cause of liver fibrosis was primarily NAFLD, the obesity and dysglycemia are the main risk factors independently optimization of the associated metabolic risk factors, including associated with longitudinal LSM changes and progression to glycemic control and body weight loss, might have contributed, moderate-to advanced liver fibrosis in the whole cohort. These at least in part, to the decline in LSM. In this sense, among risk factors also have been described in cross-sectional studies as subjects that changed from one stage to another (LSM ≥ 8.0 to factors independently associated with increased LSM in the <8 kPa or LSM ≥9.2 to <9.2 kPa), we observed a decreased in general population, apart from serum glucose, HDL-C, and body weight, waist circumference and parameters of glycemic TG levels (5). We also observed that the presence of control during the follow-up period No specific pharmacological abdominal obesity increase the risk for liver fibrosis or nutritional intervention was performed during the study progression in all groups analyzed. Some studies have period. The usual clinical practice was followed. examined the longitudinal changes of LSM in NAFLD from The major strengths of the current investigation are as outpatient clinic, not in the general population (8, 14). In follows: (i) this is a population-based study; (ii) the NAFLD subjects (n = 507), Mikolasevic et al. described a prospective design; (iii) the large sample size analyzed and (iv) progression rate to significant fibrosis assessed by TE (LSM ≥7 the measurement of LSM in all participants was performed by kPa) of 17.7% during a median follow-up period of 1.8 years. In the same experienced operator to further improve the accuracy that study, obese patients had the highest risk of progression and reduce the variability of TE measurements. (14). Lallukka et al. described that 29% of NAFLD subjects (n = However, our study had some limitations that deserve 97) had increased LSM after a median follow-up period of 11.3 mentioning: (i) Data from the present study revealed that years, with baseline liver fat (measured by proton magnetic changes in LSM values were relatively small during the follow- resonance spectroscopy) being an independent predictor of up period in the whole population. This finding might be increased LSM (8). We must highlight that, due to the explained by several reasons. First, the study was conducted in correlation between the potential risk factors associated with a community setting and not in hospital-based clinics or an LSM increase, only a few factors achieved statistical reference Hepatology units, where subjects are usually referred significance. This was more evident in models where for known liver disease. Second, we excluded subjects with abdominal obesity was found to be an independent factor for previous liver disease, including hepatitis viral infection or the increase of LSM. Both risk factors, BMI and abdominal high-risk alcohol consumption. Third, it is probable that a obesity, could not be included in the model due to high longer follow-up time is needed to identify changes in LSM in collinearity. The model with the best statistical adjustment was a general population; (ii) the XL probe was not available in this selected, but this does not mean that the excluded variable had study, but the percentage of unreliable measurements of liver no effect. stiffness was only 1.5% as our population was not overtly obese; Regarding the lipid profile, AD was a factor independently actually, few patients had class 2 and 3 obesity. The use of an XL associated with progression to moderate-to-advanced liver probe could reduce this failure rate, but we do not believe it fibrosis in the whole population. High levels of TG and/or low would have impacted our main results”; (iii) finally, data on drug HDL-C have been reported to be associated with higher LSM in treatment of the study subjects was not available. some other population-based studies (5, 21). In our previous report, AD dyslipidemia was independently associated with moderate-to-advanced liver fibrosis in subjects with T2D and Conclusions NAFLD but not in those without this condition. Furthermore, we also found female sex as a protective factor against the We hereby report for the first time that minimal changes in development liver fibrosis in general cohort. This sexual LSM values occurred in a Mediterranean population without dimorphism in liver fibrosis risk has also been previously known liver disease over a median of 4.2 years follow-up time. In described for other cardiovascular risk factors, and one a general population, metabolic risk factors such as abdominal hypothesis is that a protective role of estrogens may exist (22). obesity, dysglycemia and atherogenic dyslipidemia are Highlight in our study, we found that 64% of subjects with associated with a risk of progression to moderate-to-advanced LSM ≥8.0 kPa at baseline showed a decline <8 kPa at follow-up. liver. Specifically, abdominal obesity is a common risk factor for Similarly, results occur in NAFLD and dysglycemia subgroups. liver fibrosis progression in all groups evaluated. Of note, in a Frontiers in Endocrinology 08 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 non-negligible percentage of subjects in our cohort, a decline in 20 program, H20/20-SC1-2016- RTD, Institució Catalana de LSM was observed. Weight loss and control of other associated Recerca I Estudis Avançats Academy Award (P G). metabolic factors may have contributed to the regression of liver fibrosis at follow-up. Acknowledgments Data availability statement The authors are indebted to Echosens (Paris, France) for providing the Fibroscan system used in the current study The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Conflict of interest Ethics statement Author JF-N has received advisory and or speaking fees from Astra-Zeneca, Ascensia, Boehringer Ingelheim, GSK, Lilly, MSD, The studies involving human participants were reviewed and Novartis, NovoNordisk and Sanofi; he has received research approved by Ethics Committee of the Fundació Gol i Gorina grants to the institution from Astra-Zeneca, GSK, Lilly, MSD, (P11/58) (Barcelona, Spain). The patients/participants provided Novartis, NovoNordisk, Sanofi and Boehringer. Author KC has their written informed consent to participate in this study. received research support as principal investigator from National Institutes of Health NIH, Cirius, Inventiva, Janssen, Novartis, Novo Nordisk, and Zydus. He is consultant for Author contributions Allergan, AstraZeneca, BMN, Cirius, Coherus, Deuterex, Janssen, Genentech, Gilead, Merck, Novo Nordisk, Pzifer, MJ and SB interpreted the data and wrote the manuscript. Poxel. Author DM has received advisory and/or speaking fees NA, GP, LC and DM designed the study, interpreted data, and from Almirall, AstraZeneca, Boehringer Ingelheim, Lilly, MSD, reviewed the manuscript. AP-M, BS, RM, CE, MP-D, JF-N, PT, Novartis, Novo Nordisk and Sanofi. He has received research KC, JJ contributed to reviewed the manuscript. All authors grants to the institution from AstraZeneca, GSK, Lilly, MSD, contributed significantly to the work and approved the final Novartis, NovoNordisk, Sanofi and Boehringer Ingelheim. article. All authors contributed to the article and approved the The remaining authors declare that the research was conducted submitted version. in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding Publisher’s note The project received a research grant from the Carlos III Institute of Health, Ministry of Economy and Competitiveness All claims expressed in this article are solely those of the (Spain), awarded on the 2011 call under the Health Strategy authors and do not necessarily represent those of their affiliated Action 2013–2016, within the National Research Program organizations, or those of the publisher, the editors and the oriented to Societal Challenges, within the Technical, Scientific reviewers. Any product that may be evaluated in this article, or and Innovation Research National Plan 2013–2016, with claim that may be made by its manufacturer, is not guaranteed reference PI11/0267, PI14/00407, co-funded by European or endorsed by the publisher. Union European Regional Development Fund funds. Also supported by grants from Fondo de Investigació n Sanitaria Instituto de Salud Carlos III-Subdirecció n General de Supplementary material Evaluació n and the European Regional Development Fund Fondo Europeo de Desarrollo Regional (PI16/00043), Carlos The Supplementary Material for this article can be found III Institute of Health (PI21/00817), the Agencia de Gestió online at: https://www.frontiersin.org/articles/10.3389/ d’Ajuts Universitaris i de Recerca, the European Horizon 20/ fendo.2022.1051958/full#supplementary-material Frontiers in Endocrinology 09 frontiersin.org Julián et al. 10.3389/fendo.2022.1051958 References 1. Bazerbachi F, Haffar S, Wang Z, Cabezas J, Arias-Loste MT, Crespo J, et al. elastographic methods of liver steatosis and fibrosis; controlled attenuation Range of normal liver stiffness and factors associated with increased stiffness parameter and liver stiffness measurements. J Diabetes Complications [Internet] measurements in apparently healthy individuals. Clin Gastroenterol Hepatol (2019) (2020) 34(3):107512. doi: 10.1016/j.jdiacomp.2019.107512 17(1):54–64.e1. doi: 10.1016/j.cgh.2018.11.017 15. Stefan N, Cusi K. A global view of the interplay between non-alcoholic fatty 2. Roulot D, Czernichow S, Le Clé siau H, Costes JL, Vergnaud AC, Beaugrand liver disease and diabetes. Lancet Diabetes Endocrinol (2022) 10(4):284–96. doi: M. Liver stiffness values in apparently healthy subjects: Influence of gender and 10.1016/S2213-8587(22)00003-1 metabolic syndrome. J Hepatol (2008) 48(4):606–13. doi: 10.1016/ j.jhep.2007.11.020 16. Barb D, Repetto EM, Stokes ME, Shankar SS, Cusi K. Type 2 diabetes mellitus increases the risk of hepatic fibrosis in individuals with obesity and 3. Koehler EM, Plompen EPC, Schouten JNL, Hansen BE, Darwish Murad S, nonalcoholic fatty liver disease. Obesity. (2021) 29(11):1950–60. doi: 10.1002/ Taimr P, et al. Presence of diabetes mellitus and steatosis is associated with liver oby.23263 stiffness in a general population: The Rotterdam study. Hepatology. (2016) 63 (1):138–47. doi: 10.1002/hep.27981 17. Ciardullo S, Monti T, Perseghin G. High prevalence of advanced liver fibrosis assessed by transient elastography among U.S. adults with type 2 diabetes. 4. Petta S, Di Marco V, Pipitone RM, Grimaudo S, Buscemi C, Craxì A, et al. Diabetes Care (2021) 44(2):519–25. doi: 10.2337/dc20-1778 Prevalence and severity of nonalcoholic fatty liver disease by transient elastography: Genetic and metabolic risk factors in a general population. Liver 18. Lomonaco R, Leiva EG, Bril F, Shrestha S, Mansour L, Budd J, et al. Int (2018) 38(11):2060–8. doi: 10.1111/liv.13743 Advanced liver fibrosis is common in patients with type 2 diabetes followed in the outpatient setting: The need for systematic screening. Diabetes Care (2021) 44 5. Caballerı́a L, Pera G, Arteaga I, Rodrı́guez L, Alumà A, Morillas RM, et al. (2):399–406. doi: 10.2337/dc20-1997 High prevalence of liver fibrosis among European adults with unknown liver disease: A population-based study. Clin Gastroenterol Hepatol [Internet] (2018) 16 19. Portillo-Sanchez P, Bril F, Maximos M, Lomonaco R, Biernacki D, Orsak B, (7):1138–1145.e5. doi: 10.1016/j.cgh.2017.12.048 et al. High prevalence of nonalcoholic fatty liver disease in patients with type 2 diabetes mellitus and normal plasma aminotransferase levels. J Clin Endocrinol 6. Juliá n MT, Pera G, Soldevila B, Caballerı́a L, Julve J, Puig-Jové C, et al. Metab (2015) 100(6):2231–8. doi: 10.1210/jc.2015-1966 Atherogenic dyslipidemia, but not hyperglycemia, is an independent factor associated with liver fibrosis in subjects with type 2 diabetes and NAFLD: A 20. Unalp-Arida A. Transient elastography assessed hepatic steatosis and population-based study. Eur J Endocrinol (2021) 184(4):587–96. doi: 10.1530/EJE- fibrosis are associated with body composition in the united states. Clin 20-1240 Gastroenterol Hepatol (2021) 5(February):S1542–3565. doi: 10.1016/ j.cgh.2021.02.009 7. Kim MN, Kim SU, Kim BK, Park JY, Kim DY, Ahn SH, et al. Long-term changes of liver stiffness values assessed using transient elastography in patients 21. Kwok R, Choi KC, Wong GLH, Zhang Y, Chan HLY, Luk AOY, et al. with chronic hepatitis b receiving entecavir. Liver Int (2014) 34(8):1216–23. doi: Screening diabetic patients for non-alcoholic fatty liver disease with controlled 10.1111/liv.12377 attenuation parameter and liver stiffness measurements: A prospective cohort study. Gut. (2016) 65(8):1359–68. doi: 10.1136/gutjnl-2015-309265 8. Lallukka S, Sädevirta S, Kallio MT, Luukkonen PK, Zhou Y, Hakkarainen A, et al. Predictors of liver fat and stiffness in non-alcoholic fatty liver disease 22. Ballestri S, Nascimbeni F, Baldelli E, Marrazzo A, Romagnoli D, Lonardo A. (NAFLD) - an 11-year prospective study. Sci Rep (2017) 7(1):1–10. doi: 10.1038/ NAFLD as a sexual dimorphic disease: Role of gender and reproductive status in s41598-017-14706-0 the development and progression of nonalcoholic fatty liver disease and inherent cardiovascular risk. Adv Ther (2017) 34(6):1291–326. doi: 10.1007/s12325-017- 9. Fung J, Lai CL, Wong DKH, Seto WK, Hung I, Yuen MF. Significant changes 0556-1 in liver stiffness measurements in patients with chronic hepatitis b: 3-year follow- up study. J Viral Hepat (2011) 18(7):200–5. doi: 10.1111/j.1365-2893.2010.01428.x 23. Ginès P, Graupera I, Lammert F, Angeli P, Caballeria L, Krag A, et al. Screening for liver fibrosis in the general population: a call for action. Lancet 10. Bedogni G, Bellentani S, Miglioli L, Masutti F, Passalacqua M, Castiglione A, Gastroenterol Hepatol [Internet] (2016) 1(3):256–60. doi: 10.1016/S2468-1253(16) et al. The fatty liver index: A simple and accurate predictor of hepatic steatosis in the 30081-4 general population. BMC Gastroenterol (2006) 6:1–7. doi: 10.1186/1471-230X-6-33 11. Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella M, et al. 24. Romero-Gó mez M, Zelber-Sagi S, Trenell M. Treatment of NAFLD with The diagnosis and management of nonalcoholic fatty liver disease: Practice diet, physical activity and exercise. J Hepatol (2017) 67(4):829–46. doi: 10.1016/ guidance from the American association for the study of liver diseases. j.jhep.2017.05.016 Hepatology. (2018) 67(1):328–57. doi: 10.1002/hep.29367 COPYRIGHT 12. Castera L, Friedrich-Rust M, Loomba R. Noninvasive assessment of liver © 2023 Julian ́ , Ballesta, Pera, Peŕ ez-Montes de Oca, Soldevila, Caballerıa, ́ disease in patients with nonalcoholic fatty liver disease. Gastroenterol [Internet] (2019) 156(5):1264–1281.e4. doi: 10.1053/j.gastro.2018.12.036 Morillas, Expoś ito, Martınez–Escude ́ ,́ Puig-Domingo, Franch-Nadal, Toran ́ , 13. Eddowes PJ, Sasso M, Allison M, Tsochatzis E, Anstee QM, Sheridan D, Cusi, Julve, Mauricio and Alonso. This is an open-access article distributed et al. Accuracy of FibroScan controlled attenuation parameter and liver stiffness under the terms of the Creative Commons Attribution License (CC BY). measurement in assessing steatosis and fibrosis in patients with nonalcoholic fatty The use, distribution or reproduction in other forums is permitted, liver disease. Gastroenterol [Internet] (2019) 156(6):1717–30. doi: 10.1053/ provided the original author(s) and the copyright owner(s) are credited j.gastro.2019.01.042 and that the original publication in this journal is cited, in accordance 14. Mikolasevic I, Lukenda Zanko V, Jakopcic I, Domislovic V, Mijic A, with accepted academic practice. No use, distribution or reproduction Stevanovic T, et al. Prospective evaluation of non-alcoholic fatty liver disease by is permitted which does not comply with these terms. Frontiers in Endocrinology 10 frontiersin.org