Abstract
We aimed to investigate the relationship between metabolomic signatures of healthy lifestyle with incident metabolic dysfunction-associated steatotic liver disease (MASLD) risk, and quantify the extent to metabolic signatures explain the healthy lifestyle-MASLD relationship. This prospective cohort study analyzed 179,261 UK Biobank participants with available nuclear magnetic resonance metabolomics data. Healthy lifestyle scores incorporated dietary quality, alcohol consumption, physical activity, and smoking behavior. Elastic net regularized regression identified lifestyle-associated metabolic signatures from 251 metabolic biomarkers. Cox regression models evaluated associations between lifestyle indices, metabolic profiles, and incident MASLD. Counterfactual-based mediation analysis quantified mechanistic pathways. During follow-up, 2422 participants (1.35%) developed incident MASLD. We identified a 94-metabolite signature strongly reflecting healthy lifestyle behaviors, dominated by lipoprotein subclasses (61.70%) and fatty acids (10.64%). Each 1-unit increment in the metabolic signature corresponded to 65.9% reduced MASLD risk (HR = 0.341, 95% CI 0.311–0.373). Mediation analysis revealed that metabolic alterations explained 55.80% (95% CI 47.85–85.28%) of the protective lifestyle-MASLD association, with fatty acid metabolism contributing the largest mediation effect. Addtionly, the parallel mediating effect of metabolic signatures, BMI, diabetes, and hypertension reached 86.21%. This comprehensive metabolomic signature captures healthy lifestyle behaviors and strongly predicts incident MASLD risk through metabolic reprogramming, particularly fatty acid and lipoprotein metabolism.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-026-36704-x.
Keywords: Metabolomic, Healthy lifestyle, Metabolic dysfunction-associated steatotic liver disease, Lipoprotein, Fatty acids
Subject terms: Biochemistry, Biomarkers, Diseases, Endocrinology, Risk factors
Introduction
Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly known as non-alcoholic fatty liver disease (NAFLD), has emerged as one of the most pressing global health challenges1,2. The epidemiological trajectory demonstrates a concerning upward trend, with age-standardized prevalence rates increasing from 12,085.09 per 100,000 population in 1990 to 15,018.07 per 100,000 in 2021, representing a sustained annual percentage change of 0.71%3,4. Approximately 30.2% of adults worldwide affected by this condition, translating to over 1.27 billion individuals as of 20213,5. The clinical significance of MASLD extends far beyond hepatic manifestations, as this condition now accounts for 82.7% of cirrhosis cases and 8.0% of liver cancer incidents globally, positioning it as a primary driver of liver-related morbidity and mortality6. The economic ramifications are equally profound, with healthcare costs attributed to MASLD and its complications imposing substantial financial burdens2. It is critical to address this escalating epidemic through evidence-based interventions and policy initiatives.
Lifestyle factors are widely recognized as important modifiable determinants in MASLD development and progression. Multiple prospective cohort studies have consistently shown protective effects, with Grinshpan et al.7 reporting a 58% risk reduction for individuals adhering to favorable lifestyles, Zhang et al.8 demonstrating a 45% lower risk with high lifestyle scores in 13,303 Chinese adults, and Chang et al.9 observing a 19% risk reduction with more pronounced effects in younger individuals, males, and those with lower BMI. He et al.10 further demonstrated that ideal lifestyle adherence could prevent 66.8% of severe MASLD cases, while Yuan et al.11 identified causal associations between specific lifestyle components—including smoking cessation, moderate alcohol consumption, coffee intake, and physical activity—and reduced MASLD risk through Mendelian randomization analysis. Younossi et al.12 emphasized that healthy dietary patterns combined with regular exercise constitute core prevention and management strategies. However, despite well-established epidemiological associations, the underlying biological mechanisms remain incompletely understood.
Nuclear magnetic resonance (NMR) metabolomics offers a powerful tools for elucidating molecular mechanisms linking lifestyle factors to MASLD development. Comprehensive lifestyle interventions have demonstrated profound effects on circulating metabolite profiles, with studies revealing that adherence to healthy dietary patterns, regular physical activity, smoking cessation, and moderate alcohol consumption significantly alter amino acid metabolism, lipid compositions, lipoprotein distributions, and inflammatory markers13–17 Recent investigations have established compelling evidence for lifestyle-metabolite-disease pathways in cardiovascular outcomes18,19. Emerging metabolomics research has begun to elucidate the relationship between specific metabolite biomarkers and MASLD pathogenesis, with studies identifying distinct metabolomic signatures associated with hepatic steatosis, inflammation, and fibrosis progression20–22. However, whether specific healthy lifestyle-related metabolite signatures can predict future MASLD risk and mediate the lifestyle-MASLD relationship remains unexplored.
This study aimed to elucidate metabolic pathways linking healthy lifestyle behaviors with MASLD development using comprehensive NMR metabolomics data. We developed metabolic signatures reflecting adherence to healthy lifestyle through advanced elastic net regularization techniques. Subsequently, we examined whether these lifestyle-derived metabolic profiles were prospectively associated with incident MASLD risk. Additionally, we conducted mediation analyses to quantify the extent to which identified metabolic signatures explain the established healthy lifestyle-MASLD relationship.
Methods
Study design
This investigation was conducted using the UK Biobank, a large-scale prospective population-based cohort study that recruited over 500,000 participants aged 40–69 years from 22 assessment centers across England, Scotland, and Wales between 2006 and 2010. The study was established to investigate the determinants of diseases in middle-aged and older populations, with participants providing detailed baseline information, biological samples, and consent for long-term health outcome tracking through linkage to national health registries. All participants provided written informed consent, and the study received ethical approval from the North West Multi-Centre Research Ethics Committee (REC ID: 16/NW/0274). The study conducted in accordance with the principles of the Declaration of Helsinki.
For the present analysis, we focused on participants with available NMR metabolomics data (n = 274,237). We excluded individuals with incomplete NMR metabolomics measurements (n = 26,052) and those missing lifestyle data required for constructing the healthy lifestyle score (n = 67,382). Participants with a history of MASLD (n = 188) or other liver diseases at baseline were also excluded (n = 1354). These exclusion criteria yielded a final study population of 179,261 participants (sFig. 1).
Assessment of lifestyle
Based on a previous UK Biobank investigation23, we calculate the healthy lifestyle score incorporating dietary quality, alcohol consumption, physical activity, and smoking behavior. Dietary approaches to stop hypertension (DASH) was utilized to assess dietary quality emphasizing adequate intake of fruits, vegetables, whole grains, fish, dairy products, and minimizing processed foods and sugar-sweetened beverages. Alcohol consumption was evaluated as ≤ 1 drink per day for women and ≤ 2 drinks per day for men. Physical activity assessment focused on leisure-time activities including walking for pleasure, strenuous sports, and other exercises, with metabolic equivalent scores of 3.3, 8.0, and 4.0 respectively, calculating total weekly metabolic equivalent hours based on frequency and duration. Smoking status was categorized as never smokers (including those who tried smoking < 100 times lifetime) versus ever smokers. Each lifestyle component was scored as 0 (unhealthy) or 1 (healthy), yielding a composite score ranging from 0 to 4, with participants stratified into low (0–1), moderate (2), and high (3–4) healthy lifestyle adherence categories.
NMR metabolic biomarker detection
Serum metabolite concentrations were measured using a high-throughput NMR spectroscopy platform provided by Nightingale Health Ltd. (Helsinki, Finland). This standardized platform analyzed approximately 280,000 non-fasting EDTA plasma samples from UK Biobank participants, providing quantitative data for 251 metabolic biomarkers spanning lipoprotein subclasses, fatty acid compositions, amino acid metabolites, ketone bodies, and glucose metabolism intermediates. Comprehensive information regarding all measured metabolites and their biochemical groupings is presented in sTable 1.
Assessment of MALSD and other liver disease
The primary endpoint was incident MASLD, which was ascertained through hospitalization records or mortality data attributed to MASLD or MASH (metabolic dysfunction-associated steatohepatitis). Following ICD-10 classification and Expert Panel Consensus Statement guidelines, MASLD was characterized by ICD-10 codes K76.0 (fatty liver disease, not elsewhere classified) and K75.8 (other specified inflammatory liver conditions). For each participant, observation time was calculated from the date of baseline assessment to the first occurrence of MASLD diagnosis, all-cause mortality, or administrative censoring date. Other liver disease included cirrhosis, hepatocellular carcinoma, and liver-related mortality, with comprehensive details regarding all ICD-10 diagnostic codes documented in sTable 2.
Covariates measurement
Baseline covariates were systematically collected through standardized questionnaires, physical examinations, and medical record reviews. Demographic characteristics included age at recruitment, biological sex, ethnic background (categorized as White or Other), educational level (university degree versus no university degree), and annual household income (dichotomized at £18,000). Body mass index was computed using standardized measurements of height and weight obtained during baseline assessment and stratified into normal weight (< 25 kg/m2), overweight (25–29 kg/m2), and obese (≥ 30 kg/m2) categories according to WHO criteria. Medical history was ascertained through self-reported questionnaires and verified against medical records, encompassing diabetes mellitus, hypertension, cardiovascular disease (CVD, including coronary heart disease, heart failure, atrial fibrillation, and stroke), and cancer history.
Statistical analyses
Baseline characteristics underwent summarization through descriptive statistics, with continuous variables presented as mean ± standard deviation for normally distributed data or median (interquartile range) for skewed distributions, while categorical variables were expressed as frequencies and percentages. Between-group comparisons employed Student’s t-tests or Mann–Whitney U tests for continuous data and chi-square tests for categorical data. Multiple imputation by chained equations methodology addressed missing covariate information.
Construction of healthy lifestyle-associated metabolic profiles utilized elastic net regularized regression, a penalized modeling approach that integrates Ridge (L2) and LASSO (L1) penalties to manage high-dimensional metabolomic datasets efficiently. This methodology mitigates metabolite multicollinearity issues while enabling automated feature selection. Cross-validation procedures (tenfold) identified optimal tuning parameters (α and λ) through mean squared error minimization. Final metabolic profiles represented weighted linear combinations of retained metabolites with non-zero regression coefficients. Profile standardization via z-score transformation facilitated downstream analytical procedures. Spearman rank correlations quantified relationships between metabolic profiles and lifestyle indices.
Associations linking lifestyle indices, metabolic profiles, and MASLD incidence were evaluated using Cox regression modeling through three adjustment strategies. Model 1 was not adjusted for any covariate; Model 2 adjusted for demographic variables (age, sex, race, education, income); Model 3 additionally controlled for CVD and cancer history. Non-linear exposure–response patterns were explored via restricted cubic spline methodology with knot placement at the 10th, 50th, and 90th distribution percentiles. Stratified analyses examined effect modification across demographic age, sex, race, education, income, BMI, and comorbidity status subgroups, with interaction significance evaluated through multiplicative terms.
Mechanistic pathways connecting healthy lifestyle patterns to MASLD via metabolic intermediates were investigated through counterfactual-based mediation frameworks using the “CMAverse” R package. This analytical strategy partitions overall effects into direct and mediated components. Confidence interval estimation for mediation proportions employed 100 bootstrap resampling. Evalue was used to evaluate the impact of unmeasured confounding on the total effect, direct effect, and indirect effect. Meanwhile, the other potential mediating factors (BMI, diabetes and hypertension) were regarded as parallel mediating factors.
Analytical robustness was verified through three complementary sensitivity strategies: (1) complete-case analysis excluding participants with covariate missingness; (2) temporal restriction removing MASLD cases within two years post-baseline to minimize reverse causation bias; (3) population restriction to CVD-free participants at enrollment.
Statistical computations were executed in R version 4.5.0. Significance thresholds were set at two-tailed P < 0.05, with Bonferroni adjustment implemented for multiple comparison scenarios.
Results
Baseline characteristics of study participants
The study cohort included 179,261 UK Biobank participants, of whom 2,422 (1.35%) developed incident MASLD during follow-up. Participants who subsequently developed MASLD exhibited significantly higher BMI (31.16 ± 5.53 vs. 27.33 ± 4.61 kg/m2, P < 0.001), with 52.97% being obese compared to 23.23% in controls. They demonstrated lower socioeconomic status, including reduced university education (25.52% vs. 35.65%) and higher rates of low income (29.31% vs. 20.44%). Additionally, the MASLD group showed greater baseline comorbidity burden, with higher prevalences of diabetes mellitus (16.47% vs. 4.80%), hypertension (50.12% vs. 29.48%), and CVD (14.91% vs. 7.37%, all P < 0.001). Healthy lifestyle adherence was also significantly lower in those who developed MASLD (Table 1).
Table 1.
Baseline characteristics of participants stratified by incident MASLD status.
| Characteristic | Level | Overall (n = 179,261) |
Control (n = 176,839) |
MASLD (n = 2422) |
P |
|---|---|---|---|---|---|
| Age (years) | 56.25 ± 8.10 | 56.25 ± 8.10 | 56.56 ± 7.93 | 0.063 | |
| < 65 year | 147,209 (82.12) | 145,247 (82.14) | 1962 (81.01) | 0.158 | |
| ≥ 65 year | 32,052 (17.88) | 31,592 (17.86) | 460 (18.99) | ||
| Sex (%) | Female | 90,906 (50.71) | 89,721 (50.74) | 1185 (48.93) | 0.08 |
| Male | 88,355 (49.29) | 87,118 (49.26) | 1237 (51.07) | ||
| Race (%) | Other | 7116 (3.97) | 7013 (3.97) | 103 (4.25) | 0.505 |
| White | 172,145 (96.03) | 169,826 (96.03) | 2319 (95.75) | ||
| Education (%) | No university degree | 115,599 (64.49) | 113,795 (64.35) | 1804 (74.48) | < 0.001 |
| University degree | 63,662 (35.51) | 63,044 (35.65) | 618 (25.52) | ||
| Income (%) | < £18,000 | 36,854 (20.56) | 36,144 (20.44) | 710 (29.31) | < 0.001 |
| > £18,000 | 142,407 (79.44) | 140,695 (79.56) | 1712 (70.69) | ||
| BMI (kg/m2) | 27.38 ± 4.64 | 27.33 ± 4.61 | 31.16 ± 5.53 | < 0.001 | |
| < 25 kg/m2 | 58,584 (32.68) | 58,335 (32.99) | 249 (10.28) | < 0.001 | |
| 25 ~ 29 kg/m2 | 78,313 (43.69) | 77,423 (43.78) | 890 (36.75) | ||
| ≥ 30 kg/m2 | 42,364 (23.63) | 41,081 (23.23) | 1283 (52.97) | ||
| Lifestyle score | 2.24 ± 1.05 | 2.24 ± 1.05 | 2.04 ± 1.07 | < 0.001 | |
| Low | 8776 (4.90) | 8599 (4.86) | 177 (7.31) | < 0.001 | |
| Intermediate | 61,010 (34.03) | 60,174 (34.03) | 836 (34.52) | ||
| High | 109,475 (61.07) | 108,066 (61.11) | 1409 (58.18) | ||
| Smoking (%) | No | 80,260 (44.77) | 78,942 (44.64) | 1318 (54.42) | < 0.001 |
| Yes | 99,001 (55.23) | 97,897 (55.36) | 1104 (45.58) | ||
| Alcohol (%) | No | 92,694 (51.71) | 91,595 (51.80) | 1099 (45.38) | < 0.001 |
| Yes | 86,567 (48.29) | 85,244 (48.20) | 1323 (54.62) | ||
| Physical activity (%) | No | 69,917 (39.00) | 68,750 (38.88) | 1167 (48.18) | < 0.001 |
| Yes | 109,344 (61.00) | 108,089 (61.12) | 1255 (51.82) | ||
| DASH diet (%) | No | 72,992 (40.72) | 71,837 (40.62) | 1155 (47.69) | < 0.001 |
| Yes | 106,269 (59.28) | 105,002 (59.38) | 1267 (52.31) | ||
| History of diabetes mellitus (%) | No | 170,381 (95.05) | 168,358 (95.20) | 2023 (83.53) | < 0.001 |
| Yes | 8880 (4.95) | 8481 (4.80) | 399 (16.47) | ||
| History of hypertension (%) | No | 125,923 (70.25) | 124,715 (70.52) | 1208 (49.88) | < 0.001 |
| Yes | 53,338 (29.75) | 52,124 (29.48) | 1214 (50.12) | ||
| History of CVD (%) | No | 165,859 (92.52) | 163,798 (92.63) | 2061 (85.09) | < 0.001 |
| Yes | 13,402 (7.48) | 13,041 (7.37) | 361 (14.91) | ||
| History of cancer (%) | No | 163,452 (91.18) | 161,258 (91.19) | 2194 (90.59) | 0.316 |
| Yes | 15,809 (8.82) | 15,581 (8.81) | 228 (9.41) |
MASLD, metabolic dysfunction-associated steatotic liver disease; BMI, body mass index; DASH, dietary approaches to stop hypertension; CVD, cardiovascular disease.
Identification of metabolic signatures for lifestyle
Using elastic network regularized regression analysis, we successfully identified a 94-metabolite signature that reflected comprehensive metabolic responses to healthy lifestyle behaviors (Fig. 1, sFig. 2). The metabolic signature spanned diverse biochemical pathways, with relative lipoprotein lipid concentrations constituting the largest component (37.23%, n = 35), followed by lipoprotein subclasses (24.47%, n = 23). Amino acids and fatty acids each represented 10.64% (n = 10) of the signature, while glycolysis-related metabolites comprised 4.26% (n = 4) (sFig. 3). Additional components included ketone bodies (3.19%, n = 3), fluid balance markers (2.13%, n = 2), lipoprotein particle sizes (2.13%, n = 2), and other lipids (2.13%, n = 2), with single representatives from apolipoproteins, inflammation markers, and total lipids (1.06% each, n = 1).
Fig. 1.
Metabolites ranked from the highest to the lowest elastic network positive and negative regression coefficients for lifestyle.
The metabolic signature demonstrated distinctive directional relationships with healthy lifestyle adherence (Fig. 1, sTable 3). Metabolites showing the strongest positive associations with healthy lifestyle included LA_pct, Omega_3_pct, SFA, Citrate, and, Cholines. Conversely, metabolites exhibiting the most pronounced negative correlations encompassed Phosphatidylc, XXL_VLDL_FC, S_LDL_PL, S_HDL_PL_pct, and M_LDL_TG. Correlation analyses revealed robust associations between the 94-metabolite signature and both the composite lifestyle score and individual lifestyle components (smoking, alcohol, physical activity, DASH diet), as illustrated in sFig. 4. Baseline metabolite concentrations stratified by MASLD development status are presented in sTable 4.
Associations of lifestyle and the related metabolic profiles with MASLD
Strong protective associations between healthy lifestyle, related metabolic profiles, and MASLD incidence were demonstrated in Table 2. In the fully adjusted Model 3, each 1 factor increment in the lifestyle score and 1 unit increment in the metabolic signature corresponded to 14.8% (HR = 0.852, 95% CI 0.820–0.885) and 65.9% (HR = 0.341, 95% CI 0.311–0.373) reduced risk of MASLD, respectively. When participants were categorized into tertiles, those in the highest versus lowest tertile of lifestyle scores and metabolic signatures exhibited 33.2% (HR = 0.668, 95% CI 0.571–0.782) and 74.3% (HR = 0.257, 95% CI 0.220–0.301) lower MASLD risk. RCS analyses revealed linear dose–response relationships for both lifestyle scores (P-nonlinear = 0.727) and metabolic signatures (P-nonlinear = 0.105) (Fig. 2). The single healthy lifestyle factor, including smoking, physical activity and the DASH diet standard also was associated with a reduced risk of deeloping MASLD (sTable 5).
Table 2.
Associations of lifestyle score and the related metabolic profiles with MASLD.
| Exposure | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|
| HR(95%CI) | P | HR(95%CI) | P | HR(95%CI) | P | |
| Metabolite profiles of lifestyle | ||||||
| Each 1 factor increment | 0.327 (0.300,0.357) | < 0.0001 | 0.332 (0.303,0.363) | < 0.0001 | 0.341 (0.311,0.373) | < 0.0001 |
| Low | ref | ref | ref | |||
| Medium | 0.559 (0.512,0.609) | < 0.0001 | 0.566 (0.518,0.618) | < 0.0001 | 0.578 (0.530,0.631) | < 0.0001 |
| High | 0.246 (0.211,0.287) | < 0.0001 | 0.249 (0.213,0.292) | < 0.0001 | 0.257 (0.220,0.301) | < 0.0001 |
| P for trend | < 0.0001 | < 0.0001 | < 0.0001 | |||
| Lifestyle score | 0.841 (0.810,0.873) | < 0.0001 | 0.846 (0.814,0.878) | < 0.0001 | 0.852 (0.820,0.885) | < 0.0001 |
| Each 1 unit increment | ref | ref | ref | |||
| Low | 0.681 (0.579,0.801) | < 0.0001 | 0.701 (0.596,0.825) | < 0.0001 | 0.711 (0.604,0.837) | < 0.0001 |
| Medium | 0.641 (0.548,0.749) | < 0.0001 | 0.658 (0.562,0.770) | < 0.0001 | 0.668 (0.571,0.782) | < 0.0001 |
| High | < 0.0001 | < 0.0001 | < 0.0001 | |||
| P for trend | 0.327 (0.300,0.357) | < 0.0001 | 0.332 (0.303,0.363) | < 0.0001 | 0.341 (0.311,0.373) | < 0.0001 |
Model 1 was not adjusted for any covariate; age, sex, race, education and income;
Model 2 was adjusted for age, sex, race, education and income;
Model 3 was adjusted for Model 2 + history of CVD and cancer.
MASLD, metabolic dysfunction-associated steatotic liver disease; CVD, cardiovascular disease.
Fig. 2.
RCS analysis of the associations of lifestyle score and the related metabolic profiles with MASLD. Models were adjusted for age, sex, race, CVD, and cancer. MASLD: Metabolic dysfunction-associated steatotic liver disease; CVD: Cardiovascular disease.
Individual metabolite analyses (sTable 6) revealed specific biomarkers underlying the protective associations. Notable risk-reducing metabolites included XL_VLDL_CE_pct (HR = 0.523, 95% CI 0.494–0.554), PUFA_by_MUFA (HR = 0.525, 95% CI 0.503–0.548), LA_pct (HR = 0.573, 95% CI 0.552–0.596), various HDL cholesterol percentages, and so on. Conversely, several metabolites showed positive associations with MASLD risk, including L_HDL_PL_pct(HR = 1.812, 95% CI 1.703–1.927), MUFA_pct(HR = 1.727, 95% CI 1.665–1.792), M_HDL_PL_pct(HR = 1.678, 95% CI 1.620–1.737), L_VLDL_PL_pct(HR = 1.567, 95% CI 1.523–1.612), various VLDL components, and branched-chain amino acids.
Subgroup analyses (Fig. 3, sTable 7) revealed consistent protective associations across diverse population strata. Sex significantly modified the metabolite profiles -MASLD association (P-interaction = 0.009), with stronger protective effects observed in Female (HR = 0.298, 95% CI 0.259–0.344) compared to Male (HR = 0.375, 95% CI 0.331–0.425). History of CVD significantly influenced associations of metabolic signature with MASLD risk (P-interaction = 0.001), showing attenuated protective effects in participants with pre-existing CVD (HR = 0.492, 95% CI 0.38–0.636) when compared to without CVD (HR = 0.322, 95% CI 0.293–0.354). The metabolic signature maintained robust and consistent associations across other subgroups.
Fig. 3.
Subgroup of the associations of lifestyle score and the related metabolic profiles with MASLD. (A) Subgroup of the associations of metabolic profiles with MASLD; (B) Subgroup of the associations of lifestyle score with MASLD. Models were adjusted for age, sex, race, education, income, CVD, and cancer. MASLD: Metabolic dysfunction-associated steatotic liver disease; CVD: Cardiovascular disease.
Mediation of metabolic signature on the association of lifestyle with MASLD
Causal mediation analysis demonstrated that metabolic signatures substantially mediated the protective association between healthy lifestyle behaviors and MASLD risk (Table 3, sTable 8). The 94-metabolite signature served as a crucial mediator in the lifestyle-MASLD pathway, accounting for 55.80% (95% CI 47.85–85.28%) of the total protective association between lifestyle score and MASLD risk. Among individual lifestyle components, Smoking cessation and DASH diet adherence showed the strong mediation effect, with metabolic alterations explaining 37.32% and 36.98% of their protective association with MASLD risk, respectively. While physical activity exhibited a more modest but significant mediation effect of 12.48% .Addtionally, the Evalue value of total effect, direct effect, and indirect effect were all relatively high, which implied the mediation effect was robust (sTable 9). The parallel mediation analysis showed that the total mediating effect of metabolic signatures, BMI, diabetes and hypertension in the relationship between lifestyle and MASLD was as high as 86.21%.
Table 3.
Mediation proportion (95% confidence interval) of metabolite profiles, BMI, Diabetes mellitus and hypertension on the relationship between lifestyle and MASLD.
| Mediators | Exposure | |||
|---|---|---|---|---|
| Lifestyle | Smoking | Physical activity | DASH diet | |
| Total | 86.21 (67.57,100) | 43.57 (35.12,52.96) | 51.04 (41.47,57.79) | 49.51 (40.95,61.88) |
| Metabolite profiles of lifestyle | 55.8 (47.85,135.28) | 37.32 (28.48,51.57) | 12.48 (11.06,15.35) | 36.98 (25.68,44.4) |
| BMI | 27.66 (21.58,27.59) | 18.32 (14.14,18.21) | 36.76 (32.93,39.26) | 32.87 (26.86,47.47) |
| Diabetes mellitus | 1.9 (1.08,2.63) | 8.13 (5.91,6.82) | 11.62 (10.18,12.46) | 1.61 (0.73,3.14) |
| hypertension | 9.51 (5.91,9.83) | 6.89 (5.31,7.35) | 8.52 (6.9,10.2) | 1.6 (0.53,2.54) |
Models were adjusted for age, sex, race, CVD, and cancer. MASLD: Metabolic dysfunction-associated steatotic liver disease; BMI: Body mass index; DASH: Dietary approaches to stop hypertension; CVD: Cardiovascular disease.
Individual metabolite mediation analysis (sTable 10) identified specific biomarkers contributing most prominently to the overall mediation effect. Fatty acid metabolism emerged as the most influential mediator, with PUFA_by_MUFA contributing 52.62% mediation, followed by LA_pct at 51.75%. Lipoprotein metabolism markers played substantial mediating roles, with various cholesterol ester percentages contributing between 15 and 25% mediation. Additional important mediators included triglyceride-related metabolites, with M_LDL_TG contributing 18.69% and various VLDL triglyceride components mediating 5–25% of the association. These findings highlight the multi-pathway nature of lifestyle-mediated metabolic protection against MASLD development.
Sensitive analysis
Comprehensive sensitivity analyses confirmed the robustness of our findings across different analytical strategies. After excluding participants with missing covariates (n = 159,429), the metabolic signature maintained strong protective associations with MASLD (HR: 0.341, 95% CI 0.311–0.373) with mediation effects of 49.54% (sTables 10, 11). Excluding MASLD cases within two years of follow-up (n = 179,181) preserved protective associations (HR: 0.338, 95% CI 0.308–0.371) with increased mediation proportions of 58.43% (sTables 12, 13). When restricting to participants without baseline cardiovascular disease (n = 165,859), associations were slightly strengthened (HR: 0.329, 95% CI 0.301–0.360) with mediation effects reaching 61.68% (sTables 14, 15). RCS analyses consistently demonstrated linear dose–response relationships across all sensitivity analyses (sFigs. 5–7).
Discussion
This large-scale prospective cohort study successfully identified a comprehensive 94-metabolite signature capturing healthy lifestyle behaviors that strongly predicted incident MASLD risk. Each 1 factor increment in the lifestyle score and 1 unit increment in the metabolic signature demonstrated 14.8% and 65.9% reduced MASLD risk, respectively. Mediation analysis revealed that metabolic alterations accounted for 55.80% of the protective lifestyle-MASLD association, with the combined mediating effect of metabolic signatures, BMI, diabetes, and hypertension reaching 86.21%. Fatty acid metabolism and lipoprotein profiles emerged as the most influential mediating pathways. These findings provide novel mechanistic insights into how healthy lifestyle behaviors confer metabolic protection against MASLD development.
Our findings align with previous research investigating lifestyle-associated metabolomic signatures across diverse disease contexts. Jin et al.17 identified 15 lifestyle-related metabolites for chronic kidney disease risk, while Fu et al.13 demonstrated that 111 lifestyle-associated metabolites predicted coronary artery disease. Zhang et al.15 developed an 81-metabolite signature that mediated 64% of the protective association with rheumatoid arthritis, and Rios et al.16 constructed metabolite profiles comprising 24–58 metabolites that protected against type 2 diabetes and CVD. Lu et al.14 confirmed that all 44 NMR metabolites were significantly associated with lifestyle factors among diabetic individuals, with fatty acids showing consistent associations across components. Our identification of a comprehensive 94-metabolite signature represents one of the most extensive lifestyle-associated profiles characterized to date. Consistent with prior studies, our signature was dominated by lipoprotein subclasses (61.70%) and fatty acids (10.64%), reflecting the central role of lipid metabolism in lifestyle-mediated health effects. The 55.31% mediation proportion observed in our study aligns closely with previous findings, providing compelling evidence for the biological plausibility and generalizability of lifestyle-associated metabolomic signatures in chronic disease prevention.
Our findings are broadly supported by previous investigations examining plasma metabolite associations with MASLD risk and pathogenesis. Luukkonen et al.21 demonstrated that the metabolic component of MASLD was characterized by substrate surplus, including increased concentrations of glucose, glycolytic intermediates, and amino acids, which aligns with our identification of these metabolites as key components of the lifestyle-associated signature. Fotakis et al.24 revealed that increased levels of alanine, histidine, and tyrosine were associated with MASLD severity, while valine and aspartic acid demonstrated positive associations with disease progression, corresponding closely with our observed associations between branched-chain amino acids and MASLD risk. Li et al.22 identified serum glucose, lactate, glutamate/glutamine, and taurine as potential MASLD biomarkers using 1H NMR spectroscopy, supporting our finding that glycolysis-related metabolites comprised a significant portion of our protective signature. Ran et al.25 identified 65–87 metabolites as signatures of environmental exposures associated with MASLD risk, demonstrating the robustness of metabolomic approaches in capturing disease-relevant biological pathways. These convergent observations validate the biological relevance of our lifestyle-associated metabolic signature for MASLD prediction.
The mechanistic pathways linking healthy lifestyle behaviors to MASLD protection operate may through coordinated metabolic reprogramming across fatty acid metabolism, amino acid homeostasis, and insulin signaling cascades. Fatty acid composition represents the predominant mechanistic pathway, where healthy lifestyle patterns promote beneficial shifts from saturated to unsaturated fatty acids while enhancing polyunsaturated fatty acid biosynthesis. Impaired unsaturated fatty acid elongation disrupts mitochondrial function and accelerates hepatic steatosis progression, as deficient ELOVL5 activity leads to compromised cardiolipin composition and respiratory chain dysfunction26. Elevated plasma saturated and monounsaturated fatty acids increase severe MASLD risk by 85% and 74% respectively, while n-3 and n-6 polyunsaturated fatty acids confer protection through improved triglyceride metabolism and reduced inflammatory signaling27. Tissue-specific metabolic dysfunction further amplifies disease risk, with MASLD patients exhibiting reduced hepatic polyunsaturated fatty acid content alongside increased adipose tissue arachidonic acid accumulation28. Amino acid metabolism dysregulation, particularly elevated branched-chain and aromatic amino acids, reflects impaired adipose tissue catabolism and mitochondrial energy dysfunction during early hepatic fat accumulation29. Insulin resistance pathways significantly modulate these metabolic alterations, as hyperinsulinemic dietary and lifestyle patterns promote disease development through metabolomic perturbations that mediate 70% and 57% of total lifestyle effects respectively30. The integration of optimized lipid profiles, enhanced amino acid metabolism, and improved insulin sensitivity collectively establishes the metabolic foundation for lifestyle-mediated MASLD protection.
Mediation analysis revealed notable heterogeneity across lifestyle components. While smoking cessation and DASH diet adherence demonstrated robust mediation effects of 37.32% and 36.98% respectively, physical activity exhibited only 12.48% mediation despite strong protective effects against MASLD. This discrepancy likely reflects exercise’s multi-pathway protection beyond circulating metabolites. Exercise activates hepatic AMPK and downstream transcription factors that directly enhance fatty acid oxidation and mitochondrial function31. Additionally, exercise-induced myokine secretion, particularly interleukin-6 with up to 100-fold increases, exerts metabolic and anti-inflammatory effects not captured by our lipoprotein-focused signature32. Multi-matrix metabolomics studies demonstrate that most exercise-related metabolite changes occur tissue-specifically in muscle, adipose, and gut rather than circulation33. Exercise also modulates gut-liver axis through microbiota composition and bile acid metabolism34. These findings suggest that comprehensive understanding of physical activity’s MASLD protection requires integration of tissue-specific metabolic measurements, myokines, and gut microbiota assessments beyond plasma metabolomics.
The significant interaction between metabolic signature and baseline CVD status reveals a critical attenuation of protective effects in participants with pre-existing CVD. One plausible explanation is that once CVD is established, the disease pathogenesis may “switch tracks” from predominantly lipid-metabolic pathways to inflammation- and fibrosis-driven mechanisms that our metabolite signature inadequately captures35. Patients with established CVD exhibit chronic low-grade inflammation and endothelial dysfunction, which may override the metabolic perturbations reflected in our lipoprotein-rich signature. Additionally, widespread use of cardiovascular medications (statins, antihypertensives) in this high-risk population directly modulates lipid metabolism, potentially attenuating the lifestyle-metabolite-MASLD associations36. Alternatively, survival bias represents a methodological consideration, as individuals with severe metabolic dysfunction who develop CVD may experience higher mortality rates before MASLD diagnosis, leaving a selected cohort with paradoxically attenuated metabolic risk profiles37.
Our findings carry significant clinical and public health implications for MASLD prevention and management. First, these results reinforce lifestyle modification as the cornerstone of MASLD treatment, with the substantial 55.80% mediation effect providing robust metabolomic evidence for current clinical guidelines emphasizing dietary optimization, physical activity, and smoking cessation38,39. Second, specific metabolites represent potential therapeutic targets, particularly fatty acid metabolism pathways and lipoprotein profiles. These findings suggest pharmacological interventions targeting polyunsaturated fatty acid synthesis or lipoprotein metabolism could complement lifestyle modifications, offering novel therapeutic avenues for patients unable to achieve adequate lifestyle changes alone40. While the 94-metabolite signature demonstrates strong statistical associations, its clinical utility over conventional lipid panels (e.g., total cholesterol, LDL-C, HDL-C, triglycerides) for MASLD risk stratification in routine practice requires validation through comparative effectiveness studies and cost–benefit analyses.
This investigation presents several strengths. First, our study represents the largest prospective analysis examining lifestyle-associated metabolomic signatures for MASLD prediction. Second, we identified a comprehensive 94-metabolite signature reflecting healthy lifestyle behaviors. Third, our analytical approach integrated advanced methodologies including elastic net regularized regression and counterfactual-based mediation analysis. Fourth, we elucidated specific biological mechanisms linking lifestyle to MASLD protection, with fatty acid metabolism and lipoprotein profiles contributing > 37% mediation.
Several limitations warrant consideration. First, the observational design precludes definitive causal inferences, despite robust mediation analysis supporting mechanistic pathways. Second, the predominantly White European ancestry of UK Biobank participants may limit generalizability to other ethnic populations with different metabolic profiles and genetic backgrounds. Third, lifestyle assessments relied on self-reported questionnaires, potentially introducing measurement error and social desirability bias that could attenuate observed associations. Fourth, the NMR metabolomics platform, while comprehensive, captures only a subset of the human metabolome and may miss important metabolites detectable by other analytical approaches such as mass spectrometry. Fifth, MASLD diagnosis relied on hospital records and death data which may capture “new severe MASLD requiring hospitalization”, but fail to capture “new MASLD”. Moreover, the control group is not “a population without MASLD”, but may include subclinical or mild-to-moderate, non-hospitalized MASLD patients. This “double-bias” framework may distorts the research results., potentially underestimating disease prevalence and missing subclinical cases. Sixth, the 94-metabolite signature was derived and validated only within UK Biobank without external validation or assessment of selection stability across cross-validation folds, and its generalizability to populations with different genetic backgrounds and healthcare systems remains unknown. Finally, despite extensive covariate adjustment, residual confounding from unmeasured factors cannot be completely excluded.
Conclusion
In conclusion, this large-scale prospective cohort study successfully identified a comprehensive 94-metabolite signature that captures healthy lifestyle behaviors and strongly predicts incident MASLD risk in the UK Biobank cohort. Our findings demonstrate that metabolic alterations substantially mediate the protective lifestyle-MASLD association, with fatty acid metabolism and lipoprotein profiles emerging as predominant mechanistic pathways. The identified metabolomic signature may offer biological insights into MASLD pathogenesisfor clinical monitoring and therapeutic targeting. External validation in diverse populations is needed to confirm its generalizability.
Supplementary Information
Acknowledgements
We thank the faculties from UK Biobank Access Management Team for helping us in the data preparation (Application Number 170605). This work was supported by the National Natural Science Foundation of China (No. 82371597).
Author contributions
XT and BT contributed to Investigation, Funding acquisition and Writing—original draft. ZF contributed to Conceptualization, Data curation, Formal analysis and Writing–original draft. BT, SW and MH contributed to Conceptualization, Project administration, Supervision, and Writing—review and editing. All authors had the final responsibility for the decision to submit for publication.
Funding
This research was supported by the Hunan Provincial Natural Regional Joint Fund (2025JJ70502) and the Science and Technology Guidance Project of Yongzhou, China (2022-YZKJZD-023).
Data availability
The data are available from the UK Biobank, but there are restrictions on their availability. Researchers who wish to access the UK Biobank database will need to apply for access through the following link: https://www.ukbiobank.ac.uk/enable-your-research/
Declarations
Competing interests
The authors declare no competing interests.
Ethics approval and consent to participate
All participants provided written informed consent, and the study received ethical approval from the North West Multi-Centre Research Ethics Committee (REC ID: 16/NW/0274).
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Biao Tang, Email: yztangbiao@163.com.
Zhixing Fan, Email: fanzhixing@ctgu.edu.cn.
References
- 1.Miao, L., Targher, G., Byrne, C. D., Cao, Y. Y. & Zheng, M. H. Current status and future trends of the global burden of MASLD. Trends Endocrinol. Metab.35, 697–707 (2024). [DOI] [PubMed] [Google Scholar]
- 2.Younossi, Z. M. et al. Addressing the high and rising global burden of metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic dysfunction-associated steatohepatitis (Mash): From the growing prevalence to payors’ perspective. Aliment Pharmacol. Ther.61, 1467–1478 (2025). [DOI] [PubMed] [Google Scholar]
- 3.Kan, C. et al. Global burden and future trends of metabolic dysfunction-associated steatotic liver disease: 1990–2021 to 2045. Ann. Hepatol.30, 101898 (2025). [DOI] [PubMed] [Google Scholar]
- 4.Feng, G. et al. Global burden of metabolic dysfunction-associated steatotic liver disease, 2010 to 2021. JHEP Rep.7, 101271 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Amini-Salehi, E. et al. Global prevalence of nonalcoholic fatty liver disease: An updated review meta-analysis comprising a population of 78 million from 38 countries. Arch. Med. Res.55, 103043 (2024). [DOI] [PubMed] [Google Scholar]
- 6.Gu, Y. et al. The trend in incidence of non-alcoholic fatty liver disease and its impact on cirrhosis and liver cancer: An analysis from global burden of disease 2021. Public Health242, 79–86 (2025). [DOI] [PubMed] [Google Scholar]
- 7.Grinshpan, L. S. et al. A healthy lifestyle is prospectively associated with lower onset of metabolic dysfunction-associated steatotic liver disease. Hepatol Commun.8, e0583 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhang, S. et al. Adherence to a healthy lifestyle including sleep and sedentary behaviors and risk of metabolic dysfunction-associated steatotic liver disease in Chinese adults. Prev. Med.184, 107971 (2024). [DOI] [PubMed] [Google Scholar]
- 9.Chen, M. J. et al. Evidence summary of lifestyle interventions in adults with metabolic dysfunction-associated steatotic liver disease. Front Nutr.11, 1421386 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.He, P. et al. A healthy lifestyle, life’s essential 8 scores and new-onset severe NAFLD: A prospective analysis in UK biobank. Metabolism146, 155643 (2023). [DOI] [PubMed] [Google Scholar]
- 11.Yuan, S. et al. Lifestyle and metabolic factors for nonalcoholic fatty liver disease: Mendelian randomization study. Eur. J. Epidemiol.37, 723–733 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Younossi, Z. M., Zelber-Sagi, S., Henry, L. & Gerber, L. H. Lifestyle interventions in nonalcoholic fatty liver disease. Nat Rev Gastroenterol Hepatol.20, 708–722 (2023). [DOI] [PubMed] [Google Scholar]
- 13.Fu, Z. et al. Association between nmr metabolomic signatures of healthy lifestyle and incident coronary artery disease. Eur. J. Prev. Cardiol.30, 243–253 (2023). [DOI] [PubMed] [Google Scholar]
- 14.Lu, Q. et al. Healthy lifestyle, plasma metabolites, and risk of cardiovascular disease among individuals with diabetes. Atherosclerosis367, 48–55 (2023). [DOI] [PubMed] [Google Scholar]
- 15.Zhang, J. et al. Metabolic signature of healthy lifestyle and risk of rheumatoid arthritis: Observational and Mendelian randomization study. Am. J. Clin. Nutr.118, 183–193 (2023). [DOI] [PubMed] [Google Scholar]
- 16.Rios, S. et al. Plasma metabolite profiles associated with the world cancer research fund/American institute for cancer research lifestyle score and future risk of cardiovascular disease and type 2 diabetes. Cardiovasc. Diabetol.22, 252 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jin, T. et al. The association between metabolomic profiles of lifestyle and the latent phase of incident chronic kidney disease in the UK population. Sci Rep.15, 2299 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zheng, R. & Lind, L. A combined observational and Mendelian randomization investigation reveals Nmr-measured analytes to be risk factors of major cardiovascular diseases. Sci. Rep.14, 10645 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Tesorio, T., Mone, P., de Donato, A., Trimarco, V. & Santulli, G. Linking lifestyle factors to cardiovascular risk through metabolomics: Insights from a large population of diabetic patients followed-up for 11 Years. Atherosclerosis367, 37–39 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Fotakis, C. et al. Circulatory metabolite ratios as indicators of lifestyle risk factors based on a Greek Nafld case-control study. Nutrients.16, 1235 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Luukkonen, P. K. et al. Distinct contributions of metabolic dysfunction and genetic risk factors in the pathogenesis of non-alcoholic fatty liver disease. J. Hepatol.76, 526–535 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Li, H. et al. A Proton nuclear magnetic resonance metabonomics approach for biomarker discovery in nonalcoholic fatty liver disease. J. Proteome Res.10, 2797–2806 (2011). [DOI] [PubMed] [Google Scholar]
- 23.Zhang, Y. B. et al. Associations of healthy lifestyle and socioeconomic status with mortality and incident cardiovascular disease: Two prospective cohort studies. BMJ373, n604 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fotakis, C. et al. Serum metabolomic profiling unveils distinct sex-related metabolic patterns in Nafld. Front Endocrinol. Lausanne.14, 1230457 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ran, S. et al. Association of metabolic signatures of air pollution with MASLD: Observational and Mendelian randomization study. J. Hepatol.82, 560–570 (2025). [DOI] [PubMed] [Google Scholar]
- 26.Vouilloz, A. et al. Impaired unsaturated fatty acid elongation alters mitochondrial function and accelerates metabolic dysfunction-associated steatohepatitis progression. Metabolism162, 156051 (2025). [DOI] [PubMed] [Google Scholar]
- 27.Zhuang, P. et al. Circulating fatty acids and risk of severe non-alcoholic fatty liver disease in the UK biobank: A prospective cohort of 116 223 individuals. Food Funct.15, 10527–10538 (2024). [DOI] [PubMed] [Google Scholar]
- 28.Palomurto, S. et al. Metabolic dysfunction-associated steatotic liver disease alters fatty acid profiles in the liver and adipose tissue. J. Clin. Endocrinol. Metab.111(1), e23-31 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cheng, S. et al. Adipose tissue dysfunction and altered systemic amino acid metabolism are associated with non-alcoholic fatty liver disease. PLoS ONE10, e138889 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ha, J. et al. Dietary and lifestyle insulinemic potentials, plasma metabolome, and risk of diverticulitis: A prospective cohort study. Am. J. Clin. Nutr.120, 1053–1062 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Alabdul Razzak, I., Fares, A., Stine, J. G. & Trivedi, H. D. The role of exercise in steatotic liver diseases: An updated perspective. Liver Int. Off. J. Int. Assoc. Stud. Liver.45, e16220 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Pedersen, B. K. & Febbraio, M. A. Muscles, exercise and obesity: Skeletal muscle as a secretory organ. Nat. Rev. Endocrinol.8, 457–465 (2012). [DOI] [PubMed] [Google Scholar]
- 33.Babu, A. F. et al. Effects of exercise on nafld using non-targeted metabolomics in adipose tissue, plasma, urine, and stool. Sci. Rep.-UK.12, 6485 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Deng, D. et al. HIIT versus MICT in MASLD: Mechanisms mediated by gut-liver axis crosstalk, mitochondrial dynamics remodeling, and adipokine signaling attenuation. Lipids Health Dis.24, 144 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Targher, G., Byrne, C. D. & Tilg, H. MASLD: A systemic metabolic disorder with cardiovascular and malignant complications. Gut73, 691–702 (2024). [DOI] [PubMed] [Google Scholar]
- 36.Athyros, V. G. et al. Safety and efficacy of long-term statin treatment for cardiovascular events in patients with coronary heart disease and abnormal liver tests in the Greek atorvastatin and coronary heart disease evaluation (Greace) study: A post-hoc analysis. Lancet (London, Engl).376, 1916–1922 (2010). [DOI] [PubMed] [Google Scholar]
- 37.Allen, A. M., Hicks, S. B., Mara, K. C., Larson, J. J. & Therneau, T. M. The risk of incident extrahepatic cancers is higher in non-alcoholic fatty liver disease than obesity—a longitudinal cohort study. J. Hepatol.71, 1229–1236 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Younossi, Z. M. et al. Global consensus recommendations for metabolic dysfunction-associated steatotic liver disease and steatohepatitis. Gastroenterology.169, 1017-1032.e2 (2025). [DOI] [PubMed] [Google Scholar]
- 39.Lv, H. & Liu, Y. Management of non-alcoholic fatty liver disease: Lifestyle changes. World J. Gastroenterol.30, 2829–2833 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Sun, J. et al. Research letter: The global clinical trial landscape for non-alcoholic fatty liver disease (NAFLD) current status and future prospects. Int. J. Surg.111, 6430–6434 (2025). [DOI] [PMC free article] [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 data are available from the UK Biobank, but there are restrictions on their availability. Researchers who wish to access the UK Biobank database will need to apply for access through the following link: https://www.ukbiobank.ac.uk/enable-your-research/