Abstract
The present study aims to investigate the relationship between blood selenium and NAFLD. Data were obtained from the National Health and Nutrition Examination Survey (2017–2020), a total of 3940 eligible participants were included in this study. Multivariable logistic regression, subgroup analyses, and smooth curve fitting assessed the blood selenium-NAFLD relationship. A piecewise linear regression model identified potential thresholds. Model robustness was evaluated using multiple imputation. Among 3,940 participants, NAFLD prevalence was 45% (n = 1,173). Compared to the lowest blood selenium quartile (Q1: 103.10–169.49 µg/L), the adjusted odds ratios for NAFLD were 1.44 (95% CI: 1.16–1.78; P < 0.01) in Q3 (184.56–201.29 µg/L) and 1.26 (95% CI: 1.02–1.57; P = 0.035) in Q4 (201.30–562.23 µg/L) after full covariate adjustment (Q2 was non-significant). Each 1-standard deviation increase in log-transformed blood selenium corresponded to an adjusted odds ratios of 1.16 (95% CI: 1.08–1.26) for NAFLD. Gender significantly modified this association (P for interaction < 0.05). Adjusted smooth curve fitting demonstrated a significant non-linear positive dose-response relationship (P for non-linearity = 0.026). Elevated blood selenium concentration is significantly associated with an increased risk of NAFLD in US adults, exhibiting a non-linear dose-response pattern. This finding requires confirmation in future prospective cohort studies. Such association, if confirmed, will be of considerable public health relevance given the epidemic of NAFLD.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10238-025-01984-6.
Keywords: Non-alcoholic fatty liver disease, Blood selenium level, NHANES
Introduction
Non-alcoholic fatty liver disease (NAFLD) is the most prevalent chronic liver disease worldwide and one of the primary causes of severe liver disease, affecting approximately 25% of the global population [1–3]. NAFLD is defined as excessive fat infiltration into the liver in the absence of substantial alcohol intake or secondary causes [4]. Epidemiologic studies have suggested that increased caloric intake and a sedentary lifestyle are important risk factors of NAFLD [5]. However, it should be noted that the potential contributions of trace elements to the increasing prevalence of NAFLD have received less attention.
Selenium, an essential trace mineral for human health, is absorbed through the intestinal lumen and exerts its biological functions primarily through incorporation into human proteins as selenocysteine [6]. Selenium plays an important role in redox homeostasis, thyroid hormone metabolism, and defense against oxidative stress and inflammation [7]. Several studies showed diets with selenium supplement could increase the activity of selenoproteins in the liver and improved liver steatosis, injury and fibrosis in NAFLD mice models [8]. However, only a few epidemiologic studies have explored the associations between selenium and NAFLD and yield inconsistent findings [9–12]. The potential roles of selenium exposure in NAFLD pathogenesis remain incompletely understood.
Therefore, the objective of this study was to investigate the associations between serum selenium concentrations and NAFLD prevalence using cross-sectional data from the 2017–2020 National Health and Nutrition Examination Survey (NHANES).
Methods
Data sources
The data was obtained from the NHANES database (2017–2020). NHANES, a multistage and complex study, includes demographic, socioeconomic, dietary, health-related questionnaires, and examination data, conducted by National Centre for Health Statistics (NCHS). Among 15,560 participants in NHANES 2017–2020, we firstly selected 9023 participants who completed the elastography exam. Then, we excluded those participants in the following order: (1) with viral hepatitis B or C positive; (2) heavy drinkers [consumed more than two (female) or three (male) standard alcoholic drinks per day on average]; (3) with missing data of marital status, education level status, physical activity status, smoking status, the ratio of family income to poverty(PIR), body mass index (BMI), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), blood/dietary selenium. Finally, 3940 participants were included in this study (Fig. 1). NHANES 2017–2020 was approval by The NCHS Ethics Review Board, and inform consent was obtained from all participants. The acquisition and analysis of data was consistent with NHANES research requirements.
Fig. 1.
Study flowchart
Study variables
The exposure variable in this study was blood selenium. Concentration of blood selenium was measured by Inductively Coupled Plasma Mass Spectrometer with Dynamic Reaction Cell Technology (ELAN® DRC II, PerkinElmer Norwalk) using whole blood specimens after a dilution sample preparation step. The lower limit of detection was 24.48 µg/L. All of data were above the lower limit of detection.
There were also many confounding variables, including continuous and categorical variables, that might affect the relationship between the independent variable blood selenium and the dependent variable NAFLD status. Continuous variables included age, BMI, PIR, TC, HDL-C, dietary selenium intake. BMI was calculated by dividing kilograms by weight in meters squared. Categorical variables included gender, race/ethnicity, education level status, marital status, physical activity status, diabetes, hypertension, smoking status. Type 2 diabetes mellitus was diagnosed when ≥ 1 of the following criteria were met: (1) Self-reported diabetes diagnosis; (2) Current use of glucose-lowering medication; (3) Fasting plasma glucose ≥ 126 mg/dL (7.0 mmol/L); (4) Random plasma glucose ≥ 200 mg/dL (11.1 mmol/L); (5) Glycohemoglobin ≥ 6.5% [13]. Hypertension was defined as systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90mmHg or self-reported use of antihypertensive medications [14]. Smoking status was divided into three types: never smokers were those who responded “NO” to the question: “Smoked at least 100 cigarettes in life (%)”; current smokers were those who responded “YES” to the above question and the question: “SMQ040 - Do you now smoke cigarettes?”; ever smokers were those who responded “YES” to the question “Smoked at least 100 cigarettes in life (%)” and but “NO” to the question “SMQ040 - Do you now smoke cigarettes?” The amount of daily alcohol consumption was also collected in two 24-h recalls. If an individual completed both 24-h recalls, we used the average alcohol intake from the two 24-h recalls. Otherwise, we used the data from the first 24-h recall. Significant alcohol consumption is confirmed as > 2 drinks per day for women and > 3 drinks per day for men in the last 12 months [15]. Educational attainments, including less than high school graduate, high school graduate or GED, some college, or college graduate or above, were extracted from the self-reported questionnaire data [16]. If participants did moderate/vigorous work/recreational activities, they were identified as having physical activity. Physical activity status was divided into two types: moderate activity were those who responded “YES” to the questionnaires PAQ605, PAQ620, PAQ650, PAQ665, and the rest are defined as no moderate activity. The data of dietary selenium intake was obtained from the dietary interview component, called What We Eat in America. Dietary selenium intake is defined as the average selenium intake over a 48-hour period from dietary interview questionnaires. The concentration of dietary selenium intake did not include the nutrient from supplements and medications. The details of data can be publicity obtained at http://www.cdc.gov/nchs/nhanes/.
Definition of NAFLD
Liver stiffness and controlled attenuation parameter (CAP, a novel physical parameter related to liver steatosis) were detected by vibration-controlled transient elastography (VCTE) using FibroScan model 502 V2 Touch (Echosens, North America) with a medium (M) or extra-large (XL) wand (probe) in the NHANES Mobile Examination Centre. According to a recent landmark study, CAP values, also known as CAP ≥ 274 dB/m was considered indicative of NAFLD status because of 90% sensitivity in detecting all degrees of hepatic steatosis [17]. NAFLD was diagnosed as the presence of hepatic steatosis in the absence of excessive alcohol consumption or other chronic liver diseases [18, 19].
Statistical analysis
Participant characteristics are expressed as means and corresponding 95% CIs for continuous variables and percentages and 95% CIs for categorical variables. One way analysis of variance was used for normally distributed data, and the Kruskal-Walli’s test was used for data with a skewed distribution. Categorical variables are expressed as proportion (%), and continuous variables are expressed as mean (SD) or median [interquartile range (IQR)], as appropriate. Differences between groups were assessed using one-way analysis of variance (for normally distributed data), the Kruskal-Walli’s test for data with a skewed distribution, and χ2 tests for categorical variables. Multivariate logistic regression models were employed to estimate the odds ratio (OR) and the corresponding 95% confidence interval (CI) for the prevalence of NAFLD based on quartiles of blood selenium concentrations. Quartile-transformed blood selenium concentrations served as rank variables to explore the linear trend effect. When blood selenium was regarded as a continuous variable, OR and its 95% CI were reported for each 1SD increment in log-transformed blood selenium concentrations. No covariates were adjusted in Model 1. Model 2 was adjusted for demographic characteristics, including age, gender, race/ethnicity. Model 3 was further adjusted for age, gender, race/ethnicity, education level, marital status, physical activity status, PIR, BMI, TC, HDL-C, smoking status, dietary selenium intake, diabetes, hypertension. Subgroup analyses were performed to evaluate the relationship between the blood selenium and NAFLD in different subgroups, including age, gender, diabetes and hypertension. Smoothing curve fitting and threshold effects evaluation were applied to assess the non-linear association between blood selenium levels and the risk of developing NAFLD. We employed the likelihood ratio test to determine whether the relationship was non-linear, with P for non-linearity ≤ 0.05 indicated a non-linear relationship. We conducted sensitivity analyses by adjusting for different covariates to test the robustness of our results. The analysis was mainly performed by EmpowerStats software (http://www.empowerstats.com/cn/, X&Y solutions, Inc., Boston, MA). All probability values were two-sided, and a P-value for interaction < 0.05 was considered statistically significant.
Results
Study population
From 15,560 NHANES 2017–2020 participants aged ≥ 20 years, we excluded individuals for the following reasons: unavailable VCTE data (n = 6,537), positive viral hepatitis B/C serology or excessive alcohol consumption (n = 1,482), refused/unknown education status (n = 1,563), refused/unknown marital status (n = 4), missing data on physical activity, smoking status, BMI, or PIR (n = 853), and missing TC, HDL-C, or blood/dietary selenium measurements (n = 1,181). Consequently, 3,940 subjects comprised the final analytical cohort (Fig. 1).
Baseline characteristics based on NAFLD
In this study, 3940 participants were enrolled in main analysis. They were stratified into four groups according to quartile the concentration of blood selenium (Fig. 1). Among these, participants were attributed to four group: Q1 group (103.10–169.10.49ug/L), Q2 group (169.50–184.50.55ug/L), Q3 group (184.56–201.56.29ug/L) and Q4 group (201.30–562.30.23ug/L). A total of 1773 participants (45%) had NAFLD, and 2167 participants (55%) did not have NAFLD. Table 1 summarizes the baseline characteristics of the 3940 study subjects stratified according to the blood selenium range. The mean age (± SD) of the study subjects was 51.91 ± 17.02 years, of which 1842 (46.75%) were male. In the Q4 group (201.30–562.23.30.23 ug/L), men had higher blood selenium than women [n = 508 (51.57%) vs. n = 477 (48.43%)]. Table 1 shown the baseline characteristics of participants based on four blood selenium strata. Among four groups, gender, race, education level, marital status, smoking status, diabetes, PIR, dietary selenium intake, TC, HDL-C and median CAP were all significant difference. Obviously, with higher concentration of blood selenium, there were more male participants and participants had higher dietary selenium intake and HDL-C (Table 1).
Table 1.
Baseline characteristics of participants
| Characteristics | Overall | Q1 (< 103.10–169.49 µg/L) |
Q2 (169.50–184.55 µg/L) |
Q3 (184.56–201.29 µg/L) |
Q4 (201.30–562.23 µg/L) |
P-value |
|---|---|---|---|---|---|---|
| N = 3940 | N = 985 | N = 985 | N = 985 | N = 985 | ||
| Age, (y), mean(SD) | 51.91(17.02) | 48.99 (47.83 50.18) | 48.01 (46.89 49.15) | 48.41 (47.31 49.53) | 49.69 (48.62 50.79) | 0.232 |
| Gender, n, (%) | < 0.001 | |||||
| Male | 1842 (46.75) | 390 (39.59) | 457 (46.40) | 487 (49.44) | 508 (51.57) | |
| Female | 2098 (53.25) | 595 (60.41) | 528 (53.60) | 498 (50.56) | 477 (48.43) | |
| Race, n, (%) | < 0.001 | |||||
| Mexican American | 402 (10.20) | 78 (7.9) | 100 (10.15) | 114 (11.57) | 110 (11.17) | |
| Non-Hispanic White | 1492 (37.87) | 360 (36.55) | 363 (36.85) | 397 (40.30) | 372 (37.77) | |
| Non-Hispanic Black | 1035 (26.27) | 330 (33.50) | 289 (29.34) | 212 (21.52) | 204 (20.71) | |
| Other Race | 1011 (25.66) | 217 (22.0%) | 233 (23.65) | 262 (26.60) | 299 (30.36) | |
| Education level, n, (%) | 0.011 | |||||
| Lower than high school | 563 (14.29) | 156 (15.84) | 141 (14.31) | 142 (14.42) | 124 (12.59) | |
| High school | 850 (21.57) | 240 (24.37) | 222 (22.54) | 187 (18.98) | 201 (20.41) | |
| College or above | 2527 (64.14) | 589 (59.80) | 622 (63.15) | 656 (66.60) | 660 (67.01) | |
| Marital status, n, (%) | < 0.001 | |||||
| Married/Living with Partner | 2407 (61.09) | 554 (56.24) | 602 (61.12) | 602 (61.12) | 649 (65.89) | |
| Widowed/Divorced/Separated | 867 (22.01) | 245 (24.87) | 202 (20.51) | 226 (22.94) | 194 (19.70) | |
| Never married | 666 (16.90) | 186 (18.88) | 181 (18.38) | 157 (15.94) | 142 (14.42) | |
| Physical activity, n, (%) | 0.060 | |||||
| No | 1104 (28.02) | 299 (30.36) | 290 (29.44) | 252 (25.58) | 263 (26.70) | |
| Yes | 2836 (71.98) | 686 (69.64) | 695 (70.56) | 733 (74.42) | 722 (73.30) | |
| Smoking status, n, (%) | 0.025 | |||||
| Current | 494 (12.54) | 139 (14.11) | 139 (14.11) | 119 (12.08) | 97 (9.85) | |
| Ever | 989 (25.10) | 244 (24.77) | 235 (23.86) | 236 (23.96) | 274 (27.82) | |
| Never | 2457 (62.36) | 602 (61.12) | 611 (62.03) | 630 (63.96) | 614 (62.34) | |
| Diabetes, n, (%) | 0.027 | |||||
| No | 3099 (78.65) | 784 (79.59) | 787 (79.90) | 787 (79.90) | 741 (75.23) | |
| Yes | 841 (21.35) | 201 (20.41) | 198 (20.10) | 198 (20.10) | 244 (24.77) | |
| Hypertension, n, (%) | 0.356 | |||||
| No | 2119 (53.78) | 526 (53.40) | 534 (54.21) | 549 (55.74) | 510 (51.78) | |
| Yes | 1821 (46.22) | 459 (46.60) | 451 (45.79) | 436 (44.26) | 475 (48.22) | |
| NAFLD | < 0.001 | |||||
| No | 2167 (55.00) | 591 (60.00) | 580 (58.88) | 500 (50.76) | 496 (50.36) | |
| Yes | 1773 (45.00) | 394 (40.00) | 405 (41.12) | 485 (49.24) | 489 (49.64) | |
| PIR | 2.80(1.63) | 1.94 (1.83 2.05) | 2.25 (2.14 2.36) | 2.20 (2.08 2.32) | 2.36 (2.24 2.48) | < 0.001 |
| BMI (kg/m2) | 29.94(7.16) | 29.31 (28.88 29.74) | 29.13 (28.71 29.55) | 29.05 (28.66 29.44) | 29.19 (28.79 29.59) | 0.916 |
| Dietary selenium intake (ug,) | 107.22(50.06) | 92.42 (89.85 95.07) | 94.98 (92.20 97.85) | 98.91 (96.05 101.85) | 100.41 (97.50 103.40) | < 0.001 |
| TC (mg/dL,) | 4.83(1.04) | 4.48 (4.43 4.54) | 4.71 (4.64 4.77) | 4.77 (4.71 4.84) | 4.92 (4.86 4.99) | < 0.001 |
| HDL-C (mmol/L) | 1.38(0.40) | 1.35 (1.33 1.37) | 1.34 (1.32 1.36) | 1.32 (1.30 1.34) | 1.31 (1.29 1.33) | 0.022 |
|
Median stiffness E (KPa) |
5.87(5.02) | 5.29 (5.14 5.43) | 5.21 (5.08 5.35) | 5.13 (5.01 5.26) | 5.28 (5.15 5.41) | 0.290 |
|
Median CAP (dB/m) |
266.10(62.06) | 251.52 (247.63 255.47) | 250.76 (246.81 254.77) | 265.31 (261.39 269.29) | 266.62 (262.62 270.67) | < 0.001 |
The continuous data were shown as mean (SD), and differences between groups were compared using a t test, one-way analyses of variance (normal distribution), and Kruskal-Wallis tests (skewed distribution). The categorical data were shown as numbers and percentages [n (%)], and differences between groups were compared using the χ2 test
BMI body mass index; PIR: ratio of family income to poverty, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol, CAP: Controlled Attenuation Parameter
Multivariate logistics regression analyses
Multiple logistic regression models were constructed to explore whether the concentration of blood selenium was independent associated with NAFLD diagnosed by VCTE (Table 2). In model 1, no variables were adjusted. Compared to Q1 group, high blood selenium level (Q3 and Q4 group) had significantly positive association with NAFLD (OR = 1.45, 95% CI: 1.22, 1.74 and 1.48, 95% CI:1.24, 1.77, P for trend < 0.001). For each 1 standard deviation increase in log-transformed blood selenium concentrations, the OR (95% CI) for NAFLD was 1.20 (1.13, 1.28). In model 2, age, gender and race/ethnicity were further adjusted. Compared to referent group, high blood selenium (Q3 and Q4 group) remained significantly positive association with NAFLD (OR = 1.38, 95% CI: 1.15, 1.65 and 1.37, 95% CI:1.14, 1.64, P for trend < 0.001). For each 1 standard deviation increase in log-transformed blood selenium concentrations, the OR (95% CI) for NAFLD was 1.16 (1.09, 1.24). To further exclude the influence of covariables, age, gender, race/ethnicity, education level, marital status, physical activity status, PIR, BMI, TC, HDL-C, smoking status, dietary selenium intake, diabetes and hypertension were adjusted in model 3. The highest blood selenium level still had significant positive association with NAFLD (OR = 1.26, 95% CI: 1.02, 1.57), with a significant positive trend observed even after adjusting for all potential confounders (P for trend = 0.002). For each 1 standard deviation increase in log-transformed blood selenium concentrations, the OR (95% CI) for NAFLD was 1.16 (1.08, 1.26).
Table 2.
Associations of blood selenium levels with NAFLD
| Selenium | Q1 (< 103.10–169.49 µg/L) |
Q2 (169.50–184.55 µg/L) |
Q3 (184.56–201.29 µg/L) |
Q4 (201.30–562.23 µg/L) |
P for trend | per 1SD increment |
|---|---|---|---|---|---|---|
| N = 985 | N = 985 | N = 985 | N = 985 | |||
| Median, ug/L | 159.40 | 177.40 | 192.10 | 214.36 | ||
| Model1 | ref(1.000) | 1.05 (0.87, 1.25)0.61 | 1.45 (1.22, 1.74) < 0.01 | 1.48 (1.24, 1.77) < 0.01 | < 0.001 | 1.20 (1.13, 1.28) |
| Model2 | ref(1.000) | 1.02 (0.85, 1.23)0.82 | 1.38 (1.15, 1.65) < 0.01 | 1.37 (1.14, 1.64) < 0.01 | < 0.001 | 1.16 (1.09, 1.24) |
| Model3 | ref(1.000) | 0.99 (0.79, 1.23)0.90 | 1.44 (1.16, 1.78) < 0.01 | 1.26 (1.02, 1.57)0.035 | 0.002 | 1.16 (1.08, 1.26) |
OR(95% CI) were calculated with the multivariate logistic regression model; Bold formatting indicated statistically signifcant diferences
Model 1: no covariates were adjusted
Model 2: age, gender and race/ethnicity were adjusted
Model 3: age, gender, race/ethnicity, education level, marital status, physical activity status, PIR, BMI, TC, HDL-C, smoking status, dietary selenium intake, diabetes, hypertension were adjusted
NAFLD non-alcoholic fatty liver disease; BMI: body mass index; PIR: ratio of family income to poverty, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol
Stratified analyses
To assess the potential impact of the relationship between blood Selenium and NAFLD, we used stratified analyses to assess whether there was an effect modifier for NAFLD by the individual variables. After stratification according to age, gender, presence of diabetes and hypertension, gender were found to be effect modification in the effects of blood Selenium and NAFLD (P for interaction < 0.05) (Fig. 2).
Fig. 2.
Association between NAFLD and blood selenium in different subgroups. Adjusted for age, gender, race/ethnicity, education level, marital status, physical activity status, PIR, BMI, TC, HDL, smoking status, dietary selenium intake, diabetes and hypertension NAFLD non-alcoholic fatty liver disease; BMI: body mass index; PIR: ratio of family income to poverty, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol
Dose-response relationships between blood selenium and NAFLD
After covariate adjustment, the smoothed curve revealed a significant non-linear relationship between blood selenium and NAFLD risk (Likelihood ratio test for non-linearity = 0.008) (Fig. 3, Table S1). Given the right-skewed distribution of blood selenium concentrations, we applied natural logarithmic transformation (ln-transformed) for subsequent analyses. Figure 4 visualizes the dose-response relationship between ln-transformed blood selenium levels and the log-transformed odds ratio of NAFLD. After full adjustment for confounders (age, gender, race/ethnicity, education, marital status, physical activity, PIR, BMI, TC, HDL-C, smoking status, dietary selenium intake, diabetes, and hypertension), a significant non-linear positive association persisted (Likelihood ratio = 0.026). Using a two-piecewise linear regression model, we identified a threshold at ln(serum selenium) = 5.01 (equivalent to 195.42 µg/L). Below this threshold, the apparent effect size was 474.23; above it, the effect size decreased to 2.33. These threshold effects are detailed in Table S1.
Fig. 3.
Smoothing curve of the relationship between blood selenium level and odds ratio of NAFLD. Adjusted for age, gender, race/ethnicity, education level, marital status, physical activity status, PIR, BMI, TC, HDL, smoking status, dietary selenium intake, diabetes and hypertension NAFLD non-alcoholic fatty liver disease; BMI: body mass index; PIR: ratio of family income to poverty, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol
Fig. 4.
Smoothing curve of the relationship between loge-transformed blood selenium level and log-transformed odds ratio of non-alcoholic fatty liver (NAFLD). A. No covariates were adjusted. B. Adjusted for age, gender, race/ethnicity, education level, marital status, physical activity status, PIR, BMI, TC, HDL, smoking status, dietary selenium intake, diabetes and hypertension NAFLD non-alcoholic fatty liver disease; BMI: body mass index; PIR: ratio of family income to poverty, TC: total cholesterol, HDL-C: high-density lipoprotein cholesterol
Sensitivity analyses
Sensitivity analyses confirmed the robustness of our primary findings (Table S2). Sequentially excluding: (1) participants using selenium supplements, (2) individuals aged ≥ 80 years, and (3) those with blood selenium > 400 µg/L consistently demonstrated positive associations between blood selenium levels and NAFLD risk.
For multiple imputation analysis (Model 4): From 9,023 participants with available VCTE data, we excluded those with hepatitis B/C (n = 205), alcoholic liver disease (n = 1,277), or missing blood selenium data (n = 421), yielding 7,120 eligible participants. We performed multiple imputation (5 imputed datasets) for variables with < 20% missingness: education status (n = 1,412), marital status (n = 1,410), PIR (n = 909), TC (n = 178), HDL-C (n = 178), dietary selenium intake (n = 1,381), physical activity (n = 1,137), smoking status (n = 1,133), and BMI (n = 55). Pooled estimates using Rubin’s rules maintained a significant positive association between blood selenium levels and NAFLD risk (Table S3).
Discussion
Our study investigated the association between serum selenium level and NAFLD in U.S. adults using nationally representative data. We found non-linear associations of serum selenium with NAFLD prevalence after adjusting for sociodemographic variables, lifestyle factors, TC, HDL-C, BMI, smoking status, dietary selenium intake, diabetes and hypertension.
Only a few epidemiologic studies have evaluated the associations between selenium and NAFLD. A positive log-linear dose-response relationship was observed in a large cross-sectional study of 8,550 Chinese adults, where median plasma selenium level in this study was 213.0 µg/L, which was similar to that in our study [12]. However, in NHANES 2011–2016 elevated serum selenium levels showed a non-linear association with NAFLD, defined by serum alanine aminotransferase activity, where median plasma selenium level in this study was 130.0 µg/L, which was lower than that in our study [20]. Given the complexity of these cross-sectional findings for selenium, further studies are required to clarify the relationship between selenium exposure and liver disease, as well as potential differential roles of selenium across disease stages. In this context, mendelian randomization (MR) has been widely applied to causal inference in complex metabolic and immune-related diseases, effectively distinguishing true causal relationships from observational associations [21, 22]. Therefore, applying similar MR frameworks to NAFLD research could further elucidate whether elevated selenium levels play a causal role in disease onset and progression.
Notably, Table 1 revealed a statistically significant association between dietary selenium intake (median: 107.22 µg/day) and NAFLD prevalence. Some canonical medical guidance are suggesting people to use the selenium as a dietary supplement daily for preventing cell-damage from the free radicals [23]. Despite the nutritional benefits, excessive intake of selenium supplements can also have adverse effects on the human body. Elevated selenium levels are associated with higher blood pressure, lipid levels, type 2 diabetes, high-grade prostate cancer and neurological diseases from observational epidemiological studies and randomized clinical trials [24–28]. Recent studies have further revealed that selenium imbalance disrupts hepatic redox homeostasis, activates oxidative stress, lipid metabolic, and inflammatory pathways, and thereby promotes the metabolic vulnerability of the liver in chronic liver diseases [29–31]. These processes likely involve complex interactions among oxidative stress, selenoprotein dysfunction, and lipid-metabolic remodeling, suggesting that future studies could apply multi-pathway modeling approaches to delineate the integrated redox–metabolic–immune mechanisms underlying hepatic vulnerability [32, 33]. This suggests a possible link between dietary selenium intake and NAFLD progression, warranting further investigation into the mechanisms underlying this relationship. Longitudinal studies could provide insights into whether selenium supplementation might mitigate NAFLD risks, offering a novel preventive strategy.
Interestingly, subgroup analyses identified gender (p for interaction = 0.044) as moderators of the blood selenium-NAFLD relationship. Gender differences in the association of selenium with health effects have also been reported [34–36]. There are sex-related differences in liver outcomes reported in the literature, however the mechanisms behind the effect modification of the association between Se and liver phenotypes, by sex, remain to be determined [37]. Further studies are needed to understand sex-specific differences in the hepatotoxicity of selenium. Additionally, we further observed BMI-stratified differences despite non-significant interaction effects. Specifically, within the Q3 subgroup, serum selenium demonstrated a positive association with NAFLD among participants having BMI ≥ 25 kg/m². In the US, NAFLD often co-occurs with obesity, insulin resistance and type 2 diabetes, and the prevalence of obesity has been linked to the prevalence of NAFLD [38]. It has been clarified that dietary selenium intake can regulate the progress of NAFLD, even liver fibrosis in animal experiments and clinical trials [8]. So, we adjusted the effect of dietary selenium intake and conducted subgroup analysis. In spite of that, high dietary selenium intake still showed the positive association to NAFLD.
This study explored the association between blood selenium level and NAFLD through a large sample cross-sectional research and adjusted for potential cofounding variables to enhance validity of the results. While it still remains some limitations. Firstly, we acknowledge that liver biopsies would provide a better NAFLD assessment; however, this is not feasible for large population-based studies, VCTE-based NAFLD should be relatively specific, as a CAP value of ≥ 274 dB/m is considered to have 90% sensitivity in detecting all degrees of hepatic steatosis [17]. Secondly, potential limitations to this study include the cross-sectional study design of NHANES, which does not allow determination of temporality, and that alcohol consumption is self-reported. Additionally, in this study, no weights were calculated in multiple regression analysis and subgroup analysis. In future studies, selenium speciation and a comparison of different biomarkers will be critical to providing a better understanding of selenium exposures and associated health risks, including NAFLD.
Conclusions
In summary, the present analysis provides evidence of non-linear associations of serum selenium with NAFLD diagnosed by VCTE among US population.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We are grateful for the contributions of all staff members and participants involved in the U.S. National Health and Nutrition Examination Survey (NHANES).We appreciate the guidance of Dr. Yi Han from the Department of Gastroenterology, the Fuyang Affiliated Hospital of Anhui Medical University, Fuyang, Anhui, China, for research design and investigation.
Abbreviations
- NAFLD
Non alcoholic fatty liver disease
- NHANES
National Health and Nutrition Examination Survey
- PIR
Ratio of family income to poverty
- BMI
Body mass index
- TC
Total cholesterol
- HDL-C
High-density lipoprotein cholesterol
- CAP
Controlled attenuation parameter
- VCTE
Vibration-controlled transient elastography
Author contributions
Jinlong Chen and Xinxin Fang contributed to conceptualization and analysis; Jinlong Chen, Hanxiang Jiang, and Xinxin Fang were responsible for material preparation and data processing; Xinxin Fang performed validation and funding acquisition; Jinlong Chen and Hanxiang Jiang wrote the original draft; Xinxin Fang handled writing, review, and editing. All authors declare no conflict of interest relevant to the content of the article.
Funding
This analysis was supported by the Health research project of Anhui Province (AHWJ2023BBa20027) and University Natural Science Research Project of Anhui Province (2024AH050766).
Data availability
The details of data can be publicity obtained at http://www.cdc.gov/nchs/nhanes/.
Declarations
Ethics approval and consent to participate
As this study utilized publicly available, de-identified NHANES data, no additional institutional review board (IRB) approval was required in accordance with NIH guidelines.
Consent for publication
Not Applicable.
Competing interests
The authors declare no competing interests.
Conflict of interest
None.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jinlong Chen and Hanxiang Jiang Jinlong Chen and Hanxiang Jiang contribute equally to this work as co-first authors.
References
- 1.Díaz LA, Fuentes-López E, Ayares G, et al. The establishment of public health policies and the burden of non-alcoholic fatty liver disease in the Americas. Lancet Gastroenterol Hepatol. 2022;7(6):552–9. [DOI] [PubMed] [Google Scholar]
- 2.Ginès P, Krag A, Abraldes JG, et al. Liver cirrhosis. Lancet. 2021;398(10308):1359–76. [DOI] [PubMed] [Google Scholar]
- 3.Rich NE, Noureddin M, Kanwal F, et al. Racial and ethnic disparities in non-alcoholic fatty liver disease in the USA. Lancet Gastroenterol Hepatol. 2021;6(6):422–4. [DOI] [PubMed] [Google Scholar]
- 4.Xie R, Liu M. Relationship between non-alcoholic fatty liver disease and degree of hepatic steatosis and bone mineral density. Front Endocrinol (Lausanne). 2022;13:857110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mitra S, De A, Chowdhury A. Epidemiology of non-alcoholic and alcoholic fatty liver diseases. Transl Gastroenterol Hepatol. 2020;5:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kryukov GV, Castellano S, Novoselov SV, et al. Characterization of mammalian selenoproteomes. Science. 2003;300(5624):1439-43.
- 7.Vinceti M, Filippini T, Wise LA. Environmental selenium and human health: an update. Curr Environ Health Rep. 2018;5(4):464–85. [DOI] [PubMed] [Google Scholar]
- 8.Polyzos SA, Kountouras J, Goulas A, et al. Selenium and selenoprotein P in nonalcoholic fatty liver disease. Hormones. 2020;19(1):61–72. [DOI] [PubMed] [Google Scholar]
- 9.Longnecker MP, Taylor PR, Levander OA, et al. Selenium in diet, blood, and toenails in relation to human health in a seleniferous area. Am J Clin Nutr. 1991;53(5):1288–94. [DOI] [PubMed] [Google Scholar]
- 10.Loomba R, Filippini T, Chawla R, et al. Exposure to a high selenium environment in Punjab, India: effects on blood chemistry. Sci Total Environ. 2020;716:135347. [DOI] [PubMed] [Google Scholar]
- 11.Wu J, Zeng C, Yang Z, et al. Association between dietary selenium intake and the prevalence of nonalcoholic fatty liver disease: a cross-sectional study. J Am Coll Nutr. 2020;39(2):103–11. [DOI] [PubMed] [Google Scholar]
- 12.Yang Z, Yan C, Liu G, et al. Plasma selenium levels and nonalcoholic fatty liver disease in Chinese adults: a cross-sectional analysis. Sci Rep. 2016;6:37288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Genuth S, Alberti KG, Bennett P, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003;26(11):3160–7. [DOI] [PubMed] [Google Scholar]
- 14.Yang N, Lu Y, Cao L, et al. The association between non-alcoholic fatty liver disease and serum ferritin levels in American adults. J Clin Lab Anal. 2022;36(2):e24225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhao Y, Li H. Association of serum vitamin C with liver fibrosis in adults with nonalcoholic fatty liver disease. Scand J Gastroenterol. 2022;57(7):872–7. [DOI] [PubMed] [Google Scholar]
- 16.Singal AK, Arsalan A, Dunn W, et al. Alcohol-associated liver disease in the United States is associated with severe forms of disease among young, females and Hispanics. Aliment Pharmacol Ther. 2021;54(4):451–61. [DOI] [PubMed] [Google Scholar]
- 17.Eddowes PJ, Sasso M, Allison M, et al. Accuracy of fibroscan controlled attenuation parameter and liver stiffness measurement in assessing steatosis and fibrosis in patients with nonalcoholic fatty liver disease. Gastroenterology. 2019;156(6):1717–30. [DOI] [PubMed] [Google Scholar]
- 18.Siddiqui MS, Vuppalanchi R, Van Natta ML, et al. Vibration-Controlled transient elastography to assess fibrosis and steatosis in patients with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol. 2019;17(1):156–e632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ciardullo S, Perseghin G. Prevalence of NAFLD, MAFLD and associated advanced fibrosis in the contemporary united States population. Liver Int. 2021;41(6):1290–3. [DOI] [PubMed] [Google Scholar]
- 20.Wang X, Seo YA, Park SK. Serum selenium and non-alcoholic fatty liver disease (NAFLD) in U.S. adults: National Health and Nutrition Examination Survey (NHANES) 2011–2016. Environ Res. 2021;197:111190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Li K, Zhang C, Deng J, et al. Causal effects of gut microbiome on HIV infection: a two-sample Mendelian randomization analysis. BMC Infect Dis. 2024;24(1):280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zhang C, Deng J, Li K, et al. Causal association of monocytes with chronic kidney disease and the mediation role of frailty: a study integrating large-scale two-sample Mendelian randomization and single-cell analysis. Arch Gerontol Geriatr. 2024;123:105435. [DOI] [PubMed] [Google Scholar]
- 23.Rayman MP. Selenium and human health. Lancet. 2012;379(9822):1256–68. [DOI] [PubMed] [Google Scholar]
- 24.Kristal AR, Darke AK, Morris JS, Tangen CM, et al. Baseline selenium status and effects of selenium and vitamin e supplementation on prostate cancer risk. J Natl Cancer Inst. 2014;106(3):djt456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lippman SM, Klein EA, Goodman PJ, et al. Effect of selenium and vitamin E on risk of prostate cancer and other cancers: the selenium and vitamin E cancer prevention trial (SELECT). JAMA. 2009;301(1):39–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Christensen K, Werner M, Malecki K. Serum selenium and lipid levels: associations observed in the National health and nutrition examination survey (NHANES) 2011–2012. Environ Res. 2015;140:76–84. [DOI] [PubMed] [Google Scholar]
- 27.Laclaustra M, Navas-Acien A, Stranges S, et al. Serum selenium concentrations and hypertension in the US population. Circ Cardiovasc Qual Outcomes. 2009;2(4):369–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Naderi M, Puar P, Zonouzi-Marand M, et al. A comprehensive review on the neuropathophysiology of selenium. Sci Total Environ. 2021;767:144329. [DOI] [PubMed] [Google Scholar]
- 29.Lai G, Zhong X, Liu H, Deng J, et al. Development of a hallmark pathway-related gene signature associated with immune response for lower grade gliomas. Int J Mol Sci. 2022;23(19):11971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Himoto T, Masaki T. Current trends of essential trace elements in patients with chronic liver diseases. Nutrients. 2020;12(7):2084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Tang C, Li S, Zhang K, Li J, et al. Selenium deficiency-induced redox imbalance leads to metabolic reprogramming and inflammation in the liver. Redox Biol. 2020;36:101519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Liu H, Lai G, Shi G, et al. The influencing factors of HIV-preventive behavior based on health belief model among HIV-negative MSMs in Western china: a structural equation modeling analysis. Int J Environ Res Public Health. 2022;19(16):10185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Zhang Y, Zhang C, Li K, et al. Identification of molecular subtypes and prognostic characteristics of adrenocortical carcinoma based on unsupervised clustering. Int J Mol Sci. 2023;24(20):15465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Li J, Lo K, Shen G, et al. Gender difference in the association of serum selenium with all-cause and cardiovascular mortality. Postgrad Med. 2020;132(2):148–55. [DOI] [PubMed] [Google Scholar]
- 35.Hu H, Bi C, Lin T, et al. Sex difference in the association between plasma selenium and first stroke: a community-based nested case-control study. Biol Sex Differ. 2021;12(1):39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Arnaud J, de Lorgeril M, Akbaraly T, et al. Gender differences in copper, zinc and selenium status in diabetic-free metabolic syndrome European population - the IMMIDIET study. Nutr Metab Cardiovasc Dis. 2012;22(6):517–24. [DOI] [PubMed] [Google Scholar]
- 37.Balakrishnan M, Patel P, Dunn-Valadez S, et al. Women have a lower risk of nonalcoholic fatty liver disease but a higher risk of progression vs men: a systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2021;19(1):61-71.e15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Loomba R, Sanyal AJ. The global NAFLD epidemic. Nat Rev Gastroenterol Hepatol. 2013;10(11):686–90. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The details of data can be publicity obtained at http://www.cdc.gov/nchs/nhanes/.