We enrolled 673 cognitively unimpaired (CU) participants ≥45 years of age who underwent Aβ PET at the memory clinic in the Department of Neurology at Samsung Medical Center in Seoul, Korea, between August 2015 and November 2021. These participants comprised volunteers who applied for comprehensive dementia evaluation advertised in the local community, memory clinic, and paper; spouses of patients who visited the memory clinic; and participants with subjective cognitive decline. All participants underwent standardized neuropsychological test battery using the Seoul Neuropsychological Screening Battery 2nd edition (SNSB-II; Kang et al., 2021), brain magnetic resonance imaging (MRI), and laboratory tests, including liver function tests. All participants met the following criteria: (1) no medical history that was likely to affect cognitive function based on Christensen’s health screening criteria (Christensen et al., 1991); (2) no objective cognitive impairment in any cognitive domain on a comprehensive neuropsychological test battery (above at least −1.0 standard deviation (SD) of age-adjusted norms on any cognitive test); and (3) independence in activities of daily living. We excluded the participants with history of viral hepatitis or alcohol-liver disease, severe WMH (cap or band >10 mm and longest diameter of deep white matter lesion >25 mm), and structural lesions including cerebral infarction, intracranial hemorrhage, brain tumors, and hydrocephalus on MRI.

This study was approved by the Institutional Review Board of Samsung Medical Center approved. Written informed consent was obtained from all the participants.

Fatty liver was defined using the HSI, which includes serum alanine aminotransferase (ALT, IU/L), serum aspartate aminotransferase (AST, IU/L), body mass index (BMI, kg/m²), sex, and presence of diabetes. The HSI was calculated using the following variables: HSI = 8 × (ALT/AST ratio) + BMI (+2, if female; +2, if diabetes mellitus).

Hepatic steatosis index has been created and validated for the detection of NAFLD in large populations, not only in Asians (Lee et al., 2010; Murayama et al., 2021) but also in Europeans (Meffert et al., 2014; Khani et al., 2022). In these validation studies in Asians, the area under the receiver operator characteristic curve was greater than 0.8. According to a validation study in a Korean population (Lee et al., 2010), fatty liver was defined as an HSI > 36 for the present study.

Liver fibrosis was assessed using the Fibrosis-4 (FIB-4) index in participants with fatty liver. The FIB-4 index was calculated using the following formula: age (years) × AST (IU/L)/[platelet count (10⁹/L) × ALT (IU/L)^1/2]. The FIB-4 index correlates highly with biopsy-proven advanced liver fibrosis (Lee et al., 2021). Furthermore, previous studies have demonstrated that the FIB-4 index was closely associated with liver disease-specific mortality (Unalp-Arida and Ruhl, 2017) and cardiovascular disease (Schonmann et al., 2021). Liver fibrosis was categorized according to the following cut-off values: ≤ 1.3, low fibrosis; > 1.3 and < 2.67, intermediate fibrosis; ≥ 2.67, high fibrosis.

All the participants underwent Aβ PET [¹⁸F-florbetaben (FBB) and ¹⁸F-flutemetamol (FMM)] PET using a Discovery STe PET/CT scanner (GE Medical Systems, Milwaukee, WI, United States). For FBB or FMM PET, a 20-min emission PET scan in dynamic mode (consisting of 4 × 5 min frames) was performed 90 min after the injection of a mean dose of 311.5 MBq FBB or 197.7 MBq FMM, respectively. Three-dimensional PET images were reconstructed in a 128 × 128 × 48 matrix with 2 mm × 2 mm × 3.27 mm voxel size using the ordered-subsets expectation maximization algorithm (FBB, iteration = 4 and subset = 20; FMM, iteration = 4 and subset = 20).

Amyloid-beta uptake was defined according to Aβ PET quantification using Centiloid values. We performed a direct comparison of FBB-FMM Centiloid (dcCL) method previously developed by our group (Cho et al., 2020) to standardize the quantification of Aβ PET images obtained using different ligands. The dcCL method for FBB and FMM PET enables the transformation of the standardized uptake value ratio (SUVR) of FBB and FMM PETs to dcCL values directly without conversion to the ¹¹C-labeled Pittsburgh compound SUVR.

There are three steps for obtaining dcCL values (Cho et al., 2020): (1) preprocessing of PET images, (2) determination of the global cortical target volume of interest (CTX VOI), and (3) conversion of dcSUVR to dcCL values. First, to preprocess the Aβ PET images, PET images were co-registered to each participant’s MR image and then normalized to a T1-weighted MNI-152 template using the SPM8 unified segmentation method. We used T1-weighted MR image correction with the N3 algorithm only for intensity non-uniformities, without applying corrections to the PET images for brain atrophy or partial volume effects. Second, we used the FBB-FMM CTX VOI, defined as the area of AD-specific brain Aβ deposition in our previous study (Cho et al., 2020). Briefly, to exclude areas of aging-related brain Aβ deposition, the FBB-FMM CTX VOI was generated by comparing SUVR parametric images (with the whole cerebellum as a reference area) between 20 typical patients with Alzheimer’s disease-related cognitive impairment (AD-CTX) and 16 healthy elderly participants (EH-CTX) who underwent both FBB and FMM PET scans. To generate the FBB-FMM CTX VOI, the average EH-CTX image was subtracted from the average AD-CTX image. We then defined the FBB-FMM CTX VOI as the area of AD-related brain Aβ accumulation common to both FBB and FMM PET. Finally, the dcSUVR values of the FBB-FMM CTX VOI were converted into dcCL units using the dcCL conversion equation. The dcCL equation was derived from the FBB-FMM CTX VOI separately for FBB and FMM PET and applied to FBB and FMM dcSUVR.

To determine the participants’ dcCL cut-off-based Aβ positivity, we applied the optimal cut-off value derived using k-means cluster analysis in 527 independent samples of participants with normal cognition. The cut-off value was set at 27.08, representing the 95th percentile of the lower cluster, and the whole cerebellum was used as a reference region.

Figure 1 presents two representative cases with respect to Aβ uptake and Aβ positivity on PET scan.

All participants underwent the SNSB-II (Kang et al., 2021), which includes standardized and validated tests of various cognitive functions. The SNSB-II evaluates many cognitive factors, including verbal and visual memory, visuo-constructive function, language, praxis, components of Gerstmann syndrome (acalculia, agraphia, right/left disorientation, and finger agnosia), and frontal/executive functions. We chose to use six cognitive measures, which are representative and important neuropsychological tests, to evaluate cognitive function in five cognitive domains as follows: (1) Memory: Seoul Verbal Learning Test and Rey–Osterrieth Complex Figure Test (RCFT) delayed recall; (2) Language: Korean version of the Boston Naming Test; (3) Visuospatial function: RCFT copy; (4) Frontal executive function: Stroop Test color reading; and (5) Attention: Digit Span Test backward.

All statistical analyses were performed separately in males and females. Independent t-tests and chi-square tests were used to compare the demographic and clinical characteristics of the participants.

To investigate the association between the presence of NAFLD and Aβ deposition, we performed linear regression analyses in males and females, with the presence of NAFLD as a predictor, and quantified dcCL values as an outcome after controlling for age, education years, hypertension, diabetes, hyperlipidemia, BMI, and Apolipoprotein E4 (APOE4) genotype. To investigate the association between the severity of NAFLD and Aβ deposition, we performed linear regression analyses in males and females, with the severity of NAFLD (no NAFLD, NAFLD with low fibrosis, NAFLD with intermediate fibrosis, and NAFLD with high fibrosis) as a predictor and quantified dcCL values as an outcome, after controlling for age, education years, hypertension, diabetes, hyperlipidemia, BMI, and APOE4 genotype. We also performed a linear trend test using linear regression analysis in males and females, with the severity of NAFLD as a continuous variable, after controlling for age, education years, hypertension, diabetes, hyperlipidemia, BMI, and APOE4 genotype. To evaluate the interaction of sex and NAFLD on Aβ deposition, we performed a linear regression analysis, with NAFLD and sex together as the main effect and NAFLD × sex as an interaction effect, after controlling for age, education years, hypertension, diabetes, hyperlipidemia, BMI, and APOE4 genotype in all participants.

A sensitivity analysis using cut-off-based categorization rather than quantified dcCL values was performed to further validate the relationship between NAFLD and Aβ deposition. We used logistic regression analysis with NAFLD as a predictor and Aβ positivity as an outcome after controlling for age, education years, hypertension, diabetes, hyperlipidemia, BMI, and APOE4 genotype.

All reported p values were two-sided, and the significance level was set at 0.05. All the analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, United States).

Among the 673 participants, 410 were females and 263 were males (Table 1). Females had a lower mean age (70.7 ± 7.7 and 71.8 ± 7.2, p = 0.049), education years (10.9 ± 4.7 and 13.8 ± 3.9, p < 0.001), and BMI (23.8 ± 3.1 and 24.7 ± 2.6, p < 0.001) than males. There were no differences between males and females in the mean dcCL values (19.29 ± 30.68 and 23.73 ± 38.74, p = 0.117), the frequency of Aβ + (24.9 and 26.6%, p = 0.679), and the frequency of APOE4 genotype (25.1 and 26.6%, p = 0.732). Females had a lower frequency of hypertension (43.4 and 52.5%, p = 0.027) and diabetes (15.4 and 25.1%, p = 0.002) than males. However, the frequency of NAFLD was higher in females than in males (48 vs. 23.2%, p < 0.001).

As illustrated in Figure 2A, among females, the presence of NAFLD (β = 0.216, p < 0.001) was predictive of increased dcCL (Table 2). However, among males, the presence of NAFLD (β = 0.191, p = 0.064) was not associated with dcCL values (Table 2). There was no interaction effect between sex and the presence of NAFLD (sex × NAFLD, p = 0.109) on dcCL values.

Regarding the severity of NAFLD, among females, the presence of NAFLD with low (β = 0.254, p = 0.039), intermediate (β = 0.201, p = 0.006), and high fibrosis (β = 0.257, p = 0.027) was predictive of increased dcCL values (Table 2; Figure 3A). However, among males, there were no differences in the dcCL values between those without and with NAFLD with low (β = 0.139, p = 0.917), intermediate (β = 0.262, p = 0.077), or high fibrosis (β = 0.018, p = 0.917; Table 2).

In the linear trend test, as illustrated in Figure 3A, the dcCL values increased as the severity of NAFLD (low, intermediate, and high fibrosis) increased in females (p for trend = 0.001), which was not the case with the males (p for trend = 0.136). There was no interaction effect between sex and NAFLD severity (sex × severity of NAFLD, p = 0.073) on dcCL values.

Regarding the categorical values of Aβ, among females, presence of NAFLD [odds ratio (OR) = 2.32, p = 0.005] was independently associated with higher Aβ positivity (Figure 2B), whereas, among males, the presence of NAFLD (OR = 1.83, p = 0.165) was not associated with Aβ positivity (Figure 2B). There was no interaction effect between sex and the presence of NAFLD (sex × NAFLD, p = 0.143) on Aβ positivity. The relationship between the severity of NAFLD and Aβ positivity showed a similar trend. Aβ positivity became higher as the severity of NAFLD (low, intermediate, and high fibrosis) increased in females (p for trend = 0.014), whereas Aβ positivity was not associated with the severity of NAFLD in males (p for trend = 0.365, Figure 3B). There was no interaction effect between sex and NAFLD severity (sex × NAFLD severity, p = 0.090) on Aβ positivity.