Table 1.
Characteristics of the studies included in the systematic review and meta-analysis.
| Study | SR/MA | Sample Size | Age (Years) | Sex (Female) % | Study Type | NAFLD Diagnosis | Key Findings | |
|---|---|---|---|---|---|---|---|---|
| Non-NAFLD | NAFLD | |||||||
| Chiu L.S. et al. [8] | MA | 1972 | 384 | 52.1 ± 12 | 51.8 | Cross-sectional | Abdominal CT | Framingham Heart Study Third Generation cohort NAFLD patients had lower E/A, e′ and higher E, E/e′, LV mass. BMI was a significant mediator between liver fat and LV diastolic dysfunction parameters. |
| Aksu E et al. [18] | MA | 43 | 31 | 32 ± 4 | 100 | Case–control | Liver ultrasonography | Diastolic dysfunction indices were higher in the group of NAFLD compared to non-NAFLD (lateral E/e′ 5.8 ± 1.9 vs. 5.5 ± 2.0, p = 0.61, septal E/e′ 7.8 ± 2.2 vs. 6.5 ± 2.0, p = 0.01). Moreover, the NAFLD group showed increased parameters of left ventricle hypertrophy, left atrial size, as well as increased inter- and intra- atrial electromechanical delay. |
| Aparci et al. [19] | MA | 102 | 56 | 34.0 ± 6.7 | 34.2 | Cross-sectional | Liver ultrasonography | NAFLD patients had lower E/A and significantly greater LA diameter. No information on E/e′ was provided. |
| VanWagner L.B. et al. [20] | MA | 2442 | 271 | 50.1 ± 3.6 | 46.5 | Cross-sectional | Abdominal CT | NAFLD patients had higher LVMi, higher E/e’—findings of subclinical cardiac remodeling in systolic as well as diastolic function. |
| Bonapace S et al. [21] | MA | 18 | 32 | 64.1 ± 4.8 | 24 | Cross-sectional | Liver ultrasonography | T2DM population Diastolic dysfunction in NAFLD patients Increased LV filling pressures |
| Cassidy S et al. [22] | MA | 19 | 19 | 55 ± 15 | 42 | Case–control | 1H-magnetic resonance spectroscopy of the liver | Adults with NAFLD and T2DM demonstrate concentric remodeling with an elevated eccentricity ratio compared to controls. No data are provided on the ratio of E/e′, with similar LVMi across the two groups. |
| Chang W. et al. [23] | MA | 30 | 67 | 47.1 ± 8.9 | 33 | Cross-sectional | Liver ultrasonography | T2DM population. No significant difference among controls and mild NAFLD. LA strain values decreased in severe NAFLD group. |
| Chung G.E. et al. [24] | MA | 1990 | 1310 | 54 ± 10.2 | 36.8 | Cross-sectional | Liver ultrasonography | Increased prevalence of LV diastolic dysfunction in NAFLD groups, as defined by E/A, E/e′, septal e′, LA, and LV dimensions. Increased risk of diastolic dysfunction according to fibrosis in non-obese patient group following patient stratification according to BMI. |
| Fallo F. et al. [25] | MA | 38 | 48 | 49 ± 10 | 32 | Case–control | Liver ultrasonography | Essential hypertensives patients No data are provided regarding E/e’ ratio. Higher prevalence of diastolic dysfunction (62.5% vs. 21.1%, p < 0.001) in NAFLD compared to control subjects, as defined by E/A ratio < 1 and E-wave deceleration time > 220 ms. |
| Fotbolcu H et al. [26] | MA | 30 | 35 | 40.3 ± 6.2 | 41.5 | Case–control | Liver ultrasonography | NAFLD patients had lower E/A and e’, increased DT, IVRT, and E/e’—increased LV filling pressures and impaired diastolic function. |
| Goland S. et al. [27] | MA | 25 | 38 | 44.8 ± 6.6 | 24.9 | Case–control | Liver biopsy | NAFLD patients had altered LV geometry with pronounced thickening of IVS and PW, lower E/A, and increased DT. However, no significant differences in LV filling pressures (E/e’) were reported. |
| Ismaiel A. et al. [28] | MA | 37 | 38 | 42.1 ± 18.8 | 53.4 | Cross-sectional | Liver ultrasonography SteatoTest |
MAFLD patients had lower E/A compared to healthy controls and increased LV filling pressures as defined by E/e’. |
| Jung J.Y. et al. [29] | MA | 14,641 | 6171 | 39.7 ± 7.6 | 37.1 | Cross-sectional | Liver ultrasonography | Impaired LV relaxation in NAFLD patients, with a correlation between NAFLD severity and degree of LV remodeling and diastolic dysfunction as measured with E/e’, LV mass, LVEDV, E/A, and tissue e’ velocities. |
| Khoshbaten M et al. [30] | MA | 30 | 30 | 40 ± 7 | 40 | Case–control | Liver ultrasonography | NAFLD patients had increased LAVi compared to the controls. |
| Kim NH et al. [31] | MA | 1465 | 421 | 56.6 ± 7.3 | 62 | Cross-sectional | Computed tomography | 4 groups: with and without NAFLD, with and without MetS. No significant differences. The presence of NAFLD in subjects with MetS additively contributed to a subclinical deterioration in LV diastolic function. |
| Kocabay G et al. [32] | MA | 21 | 55 | 42.1 ± 7.3 | 43.1 | Cross-sectional | Liver biopsy | LA geometry and functional properties assessed by speckle-tracking echo. NAFLD patients had lower peak strain during atrial and ventricular systole. LA strain during ventricular systole was significantly associated with E, Em, and LAVi values. Atrial deformation parameters did not significantly differ among NAFLD groups according to liver disease severity. |
| Lai YH et al. [33] | MA | 1019 | 1142 | 48.1 ± 7.3 | 36.3 | Retrospective cohort | Liver ultrasonography | NAFLD patients with increased fibrosis had significantly elevated E/e’, LA stiffness, decreased e’, and decreased LA strain values, independent of cardiovascular disease risk factors and obesity. |
| Lee H. et al. [34] | MA | 251 | 355 | 62.7 ± 5.1 | 75.5 | Cross-sectional | Liver ultrasonography | T2DM population LV diastolic dysfunction prevalence higher in NAFLD group with increased LV mass, LA dimensions, lower E/A ratio, and longer DT. |
| Lee M et al. [35] | MA | 48 | 83 | 60 | 44.3 | Cross-sectional | Elastography | T2DM patients NAFLD group had diastolic dysfunction with increased LV filling pressures (E/e’) and LAVi. Higher degree of hepatic fibrosis independently associated with higher E/e’ ratio and decreased myocardial FDG uptake in PET |
| Lee YH et al. [36] | MA | 190 | 118 | 57.1 | 44.9 | Cross-sectional | Elastography | NAFLD patients showed increased LV wall thickness, ventricular and atrial volumes, LV diastolic dysfunction as assessed by decreased e’ and increased LV filling pressures (E/e’), and atrial systolic dysfunction with reduced atrial longitudinal strain and increased atrial stiffness |
| Mahfouz RA et al. [37] | MA | 80 | 180 | 47.6 | 46.6 | Case–control | Elastography | NAFLD patients had increased LA stiffness index values [as calculated with (E/e’)/LA global PALS ratio], interatrial septum thickness, LAVi and E/e’. Higher AF prevalence in NAFLD group, possibly related to altered LA geometry |
| Mantovani, A. et al. [38] | MA | 64 | 158 | 67.4 | 29.6 | Cross-sectional | Liver ultrasonography | T2DM outpatient population NAFLD group had echocardiographic features of diastolic dysfunction; lower e’ and increased E/e’, LVEDP, and LAVi |
| Miller A et al. [39] | MA | 133 | 49 | 68.4 ± 12.9 | 58.6 | Cross-sectional | US/MRI/CT/biopsy/ICD-9/10 | HFpEF patients 27% met NAFLD criteria with higher rates of NYHA III-IV HF symptoms and diastolic dysfunction grade ≥ 2, increased IVS thickness and LAVi. |
| Moise CG et al. [40] | MA | 80 | 79 | 31.5 ± 6.8 | 38.8 | Case–control | Liver ultrasonography | Young (15–45) adult population Hepatic steatosis was associated with lower e’ velocities, higher E/A, E/e’. Concomitant DM did not affect diastolic dysfunction parameters. |
| Peng D et al. [41] | MA | 57 | 171 | 47.8 ± 12.1 | 32 | Cross-sectional | Liver ultrasonography or transient elastography | Moderate-to-to severe steatosis patients had higher risks for left ventricle diastolic dysfunction and cardiac remodeling with higher LVMi. |
| Saluja M et al. [42] | MA | 13 | 57 | 55.7 ± 10.4 | 45.1 | Cross-sectional | Liver ultrasonography | T2DM population NAFLD patient group had decreased e’ tissue velocities, increased E/e’ ratio, and elevated LVEDP. |
| Şerban A et al. [43] | MA | 70 | 73 | 57.5 ± 3.5 | 28.9 | Case–control | Liver ultrasonography | T2DM population NAFLD patient group had lower e’ tissue velocities, higher E/e’, more severe diastolic dysfunction compared to controls. |
| Simon TG et al. [44] | MA | 51 | 14 | 48.4 ± 12.3 | 40.1 | retrospective cohort | Liver biopsy | NAFLD patients had diastolic dysfunction echocardiographic parameters such as increased LAVi and LVMi, decreased e’ tissue velocities, E, E/A, and DT. |
| L. B. VanWagner et al. [45] | MA | 1668 | 159 | 49.9 ± 3.6 | 60.6 | Prospective cohort | Computed Tomography | From CARDIA study NAFLD patients had increased LAVi, LV mass and impaired LV relaxation with elevated LV filling pressures (higher E/e’). |
| Zamirian M et al. [46] | MA | 30 | 30 | 37.6 ± 4.7 | 48.3 | Case–control | Liver biopsy/ ultrasonography |
NAFLD patients had altered LV geometry with increased diameters as well as diastolic dysfunction with lower e’ tissue velocities and higher E/e’. |
| Canada J McN et al. [47] | SR | - | 36 | 54 (48–60) | 67 | Cross-sectional | Biopsy confirmed | NASH was compared to NAFL. Diastolic function was assessed according to liver fibrosis. E/e’ during exercise increased progressively with increasing fibrosis. NASH was associated with impaired exercise capacity compared to NAFL. |
| Fudim M et al. [48] | ΜA | 842,616 | 27,919 | 74.5 ± 7.1 | 57 | Cohort study | International Classification of Diseases | Patients with (versus without) baseline NAFLD had a significantly higher risk of new-onset HF. Among HF subtypes, the association of NAFLD with downstream risk of HF was stronger for HFpEF. |
| Furuhashi M et al. [49] | SR | - | 185 | 63 ± 14 | 43 | Cross-sectional | Fatty liver index | Elevated fatty liver index is independently associated with LV diastolic dysfunction in a general population without medication. |
| Makker J et al. [50] | SR | 94 | 64 | - | - | Case–control | Computed tomography | Severe NAFLD compared to control was associated with a higher left ventricular mass after normalization for height2.7. |
| Petta S. et al. [51] | SR | - | 147 | 48 ± 12 | 36 | Cross-sectional | Biopsy confirmed | Left ventricular mass, relative wall thickness, and left atrial volume, as well as E/A ratio and diastolic dysfunction were linked to severe liver fibrosis. |
| Sonaglioni A et al. [52] | SR | - | 92 | 54 ± 11 | 50 | Cross-sectional | Liver stiffness measurement | 12.0% of the NAFLD patients were found with normal diastolic filling pattern, 7.6% showed a pseudonormal diastolic filling pattern, and no patient was diagnosed with restrictive filling pattern. Left ventricular filling pressures as expressed by the average E/e’ ratio, were in the “gray zone” of 8 to 13 (average E/e’ ratio 10.0 ± 2.9). |
| Ybarra J et al. [53] | SR | - | 151 | 38.4 ± 07 | 76 | Cross-sectional | Liver ultrasonography | Increased prevalence of LVH according to ALT levels. Lower E/A ration according to ALT levels. |
| Yoshihisa A. Et al. [54] | SR | 492 | 69.8 ± 13.9 | 50.2 | Prospective observational | Non-alcoholic fatty liver disease fibrosis score | Patients with HFpEF and NAFLD. Higher NAFLD fibrosis score is associated with higher mortality, and higher BNP levels. |
|
SR: systematic review, MA: meta-analysis, NAFLD: non-alcoholic fatty liver disease, LA: left atrial, LVMi: left ventricular mass index, T2DM: type 2 diabetes mellitus, BMI: body mass index, IVRT: isovolumic relaxation time, DT: deceleration time, IVS: interventricular septum, PW: posterior wall, LVEDV: left ventricular end-diastolic volume, MetS: metabolic syndrome, LAVi: left atrial volume index, HFpEF: heart failure with preserved ejection fraction, NYHA: New York Heart Association, LVEDP: left ventricular end-diastolic pressure, ALT: alanine transaminase, BNP: brain natriuretic peptide.