Maternal Hb and fetal growth change dynamically throughout pregnancy. We examined the associations of time-specific Hb levels and Hb trajectories with fetal biometrics and adverse birth outcomes. This prospective study included 6844 pregnant women (mean age 26·6 (sd 3·7) years) from the Tongji-Huaxi-Shuangliu Birth Cohort. Hb levels were measured at four periods: early (6-12 gestational weeks), middle (13-27), middle-late (28-32) and late pregnancy (33-37). Fetal biometrics were assessed by ultrasound from middle to late pregnancy. Birth outcomes were obtained from medical records, including small for gestational age (SGA), low birth weight (LBW) and preterm birth. Three Hb trajectories were identified: consistent decline (Trajectory 1), consistently low (Trajectory 2) and increase from middle-late pregnancy (Trajectory 3). Compared with Trajectory 1, Trajectory 3 was associated with lower estimated fetal weight (β, -0·54; 95 % CI -0·99, -0·09) and abdominal circumference (β, -0·21; 95 % CI -0·40, -0·01) in late pregnancy and higher umbilical artery resistance index across pregnancy (β, 0·65; 95 % CI 0·31, 1·00). Trajectory 3 was also associated with higher risk of LBW (OR, 1·57; 95 % CI 1·09, 2·26). In middle-late pregnancy, higher Hb (≥ 130 g/l) was associated with higher risks of LBW (OR, 2·26; 95 % CI 1·08, 4·25) and preterm birth (OR, 2·03; 95 % CI 1·12, 3·44) compared with the reference (110-129 g/l). Elevated maternal Hb from middle-late pregnancy onwards may be associated with lower fetal weight and increased risk of LBW. Dynamic monitoring of maternal Hb may facilitate targeted nutritional management in pregnant women.

China has prioritized the prevention of cardiovascular disease. Over the past decade, significant investments have been made in public health services, along with the implementation of various health promotion policy agendas. As a result, environmental cardiovascular hazards have been substantially reduced, and care coverage within the primary health care system has improved. Nevertheless, adherence to various healthy lifestyles within the population has exhibited different secular trends, meanwhile the underutilization of secondary prevention medications among vulnerable populations is concerning. The Healthy China Initiative 2030 outlines a roadmap to prevent, pre-empt, and mitigate cardiovascular disease. China, along with other developing countries facing similar challenges, should consider fostering supportive social environments for healthy behaviors, targeting vulnerable groups for lifestyle adherence and access to care, expanding the scope of services provided by primary care providers, leveraging new technologies to develop scalable tools, and integrating surveillance systems for health promotion and care.

Background: Glycaemic heterogeneity is incompletely captured by HbA1c alone, because similar glycaemic levels can arise from different organ-system perturbations and lead to divergent complication risks. We developed a routine biomarker-based framework to define glycaemic heterogeneity as both discrete metabolic subtypes and a continuous risk architecture. Methods: We included 346,474 UK Biobank participants and 24,693 Dongfeng–Tongji participants with HbA1c and routine biomarker data. Sex-stratified ridge regression derived seven system-specific fitted HbA1c measures reflecting glucose, lipid, liver, kidney, cardiovascular, inflammatory, and body composition pathways. Variance-equalised dimensionality reduction and BIRCH clustering identified metabolic subtypes, while centroid-based proximity scores quantified continuous disease-risk gradients. Subtype-associated molecular signatures were assessed using Olink proteomics. Findings: System-specific fitted HbA1c measures shared a glycaemic backbone but identified seven reproducible subtypes with distinct risk architectures. The highest-risk subtype differed by outcome: hyperglycaemic dysregulation for diabetes and retinopathy, renal dysfunction metabolism for cardio-renal susceptibility, hepatic-dominant metabolism for liver-centred risk, inflammation-activated metabolism for chronic obstructive pulmonary disease, and metabolically healthy obesity for sleep disorders. Proximity modelling revealed disease-specific risk axes beyond subtype boundaries. The Chinese cohort reproduced major glycaemic and hepatic structures while revealing a pressure-dominant subtype. Proteomics linked these architectures to immune–metabolic activation, hepatic xenobiotic metabolism, and inflammation with impaired renal transport. Interpretation: By translating routine biomarkers into organ-specific risk profiles, this framework could support scalable risk stratification and help target prevention and intervention to the dominant metabolic pathways driving future complications.

Objective:This study aims to investigate the joint associations of sarcopenia and social isolation with mortality risk. Methods:Using data from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) and the UK Biobank, sarcopenia was diagnosed according to European and Asian Working Groups for Sarcopenia criteria. Social isolation was assessed using standardized questionnaires, including questions on solitude, frequency of social activities, contact with others, and marital status (for the CLHLS only). Results:During the follow-up period, 8,249 deaths occurred in the CLHLS and 26,670 deaths in the UK Biobank groups. While no significant interaction was observed between sarcopenia and social isolation in predicting all-cause mortality in the CLHLS cohort, the association between social isolation and mortality was stronger among individuals with sarcopenia in the UK Biobank ( P-interaction = 0.03, relative risk due to interaction: 0.23, 95% confidence interval [ CI]: 0.06-0.41). Further joint analyses showed that participants with sarcopenia and high levels of social isolation had the highest mortality risk (hazard ration [ HR]: 1.99; 95% CI: [1.74-2.28] in the CLHLS and 1.69 [1.55-1.85] in the UK Biobank) compared to those without either condition. Conclusion:The combination of social isolation and sarcopenia synergistically increases the risk of mortality in middle-aged and older adults across diverse populations.

BACKGROUND:Heart failure (HF) is a severe complication of type 2 diabetes (T2D), yet the association between plasma proteins and HF in individuals with T2D remains underexplored. OBJECTIVES:This study aimed to investigate the association between plasma proteomic profiles and HF, and further evaluate whether proteomic data could enhance HF prediction beyond clinical variables, polygenic risk, and N-terminal prohormone of brain natriuretic peptide (NT-proBNP). METHODS:This cohort study included 2198 participants with T2D from the United Kingdom Biobank. Cox proportional hazards models were employed to examine associations between 2920 plasma proteins and incident HF. Clinical and protein predictors were selected using the least absolute shrinkage and selection operator method based on 10-fold cross-validation. The predictive performance of models was evaluated using Harrell's C-index, calibration slope, net reclassification improvement, integrated discrimination improvement, decision curve analysis, and calibration plots. RESULTS:During a median follow-up of 13.1 y, 298 individuals developed incident HF. A total of 455 proteins (447 positively and 8 inversely) were associated with HF, primarily involved in cell adhesion, extracellular space, signaling receptor activity, and cytokine-cytokine receptor interaction pathways. The top protein associated with increased risk of HF was whey acidic protein (WAP) 4-disulfide core domain protein 2, with a per-SD increment hazard ratio (HR) of 1.90 [95% confidence interval (CI): 1.65, 2.19)]. Conversely, the top protein inversely associated with HF risk was apolipoprotein C-I, with a per-SD increment HR of 0.75 (95% CI: 0.66, 0.85). Seventeen proteins were subsequently selected as proteomic predictors, and the resulting 17-protein risk score improved HF prediction beyond clinical variables, polygenic risk, and NT-proBNP, yielding a maximum C-index of 0.833 with an increment of 0.091. CONCLUSIONS:This study identified proteomic biomarkers for HF, highlighted potential pathways that inform biological mechanisms, and demonstrated that proteomic data enhanced HF risk prediction in individuals with T2D.

BACKGROUND:To examine serum purine metabolites as biomarkers of maternal dietary purine intake and their associations with glycaemic control and preterm birth during pregnancy. METHODS:We included 1480 pregnant Chinese women with gestational diabetes mellitus (GDM) from the Westlake Precision Birth Cohort (WeBirth) at mid-pregnancy, divided into discovery (n = 1230) and internal validation (n = 250) subcohorts. Linear regression was used to assess associations between serum metabolites, dietary purine intake, and continuous glucose monitoring (CGM) metrics. A purine score was developed and evaluated for associations with glycaemic control and preterm birth, with replication in 936 pregnant women from the Tongji-Huaxi-Shuangliu Birth Cohort (THSBC). FINDINGS:Serum purine level was associated with dietary purine intake and with the mean of daily difference (β = 0.096 [95% CI, 0.038, 0.15]). The association between higher serum purine and preterm birth risk in the WeBirth cohort (OR = 1.29 [95% CI, 1.06, 1.52]) was replicated in the THSBC cohort. Early-pregnancy serum purine also predicted GDM risk (OR = 1.38 [95% CI, 1.13, 1.63]) in the THSBC. A purine score comprising purine, adenine, 6-O-methylguanine, and uric acid showed consistent associations with glycaemic variability and preterm birth across both cohorts. INTERPRETATION:Reflecting dietary purine intake, serum purine level and purine score are associated with greater glycaemic variability and higher preterm birth risk. Monitoring dietary purine intake may improve glycaemic control in GDM and prevent pregnancy complications. FUNDING:National Key R&D Program of China; National Natural Science Foundation of China; Zhejiang Provincial Key Laboratory Construction Project; Sichuan Provincial Natural Science Foundation.

Introduction: Cardiovascular-kidney-metabolic (CKM) syndrome represents a progressive pathophysiological continuum. While early-life famine exposure has been associated with cardiometabolic outcomes, its impact on transitions across CKM stages remains unknown. Methods: We utilized data from the Kailuan Study, an ongoing, prospective cohort study in China. We included 24,190 participants with complete baseline anthropometric and biochemical data. Eligible participants had a baseline CKM syndrome stage between 0 and 3 and at least two CKM syndrome stage assessments (through Dec 31, 2022). Famine exposure was classified by birth year into fetal-exposed (1959–1961), early-childhood-exposed (1956–1958), and unexposed (1962–1964). Multistate models estimated associations between famine exposure and stage-specific transitions—CKM progression (within stages 0–3), recovery, clinical cardiovascular disease (CVD), and mortality—stratified by baseline CKM stage. Results: Over a median follow-up of 15.9 years, 2,611 participants developed CVD and 1,645 died. Among those with baseline stage 0, famine exposure was not significantly associated with adverse progression. Among those in stage 1, fetal famine was associated with higher risks of progression (hazard ratio [HR] 1.10; 95% CI, 1.00–1.21) and death after progression (HR 1.78; 95% CI, 1.01–3.14); early-childhood famine showed similar patterns: progression (HR 1.09; 95% CI, 1.00–1.18) and death after progression (HR 1.88; 95% CI, 1.12–3.14). For stage 2, fetal and early-childhood famine were associated with increased risks of progression (HR 1.42; 95% CI, 1.22–1.66 and HR 1.89; 95% CI, 1.66–2.16, respectively), clinical CVD (HR 1.19; 95% CI, 1.05–1.34 and HR 1.30; 95% CI, 1.18–1.45), and death (early-childhood HR 1.65; 95% CI, 1.42–1.92). In addition, early-childhood famine was associated with reduced probability of recovery (HR 0.79; 95% CI 0.71-0.87). No statistically significant associations were detected in participants with stage 3 at baseline (n = 196), likely due to limited power. Conclusion: Early-life famine exposure does not raise CVD or mortality risk among metabolically healthy adults, but it markedly accelerates CKM progression once metabolic dysfunction emerges. This underscores two priorities: (1) sustain metabolic health in famine exposed populations, and (2) implementing proactive interventions at the earliest CKM stages (0-1) to prevent progression to more severe stages.

Epidemiological associations between multiple metals exposure and hypertension or blood pressure remain limited and inconclusive, especially for essential metals. This investigation intended to examine the individual and collective effects of multiple metals on hypertension and blood pressure in general Chinese adults. Whole blood levels of 8 essential metals and 3 nonessential metals were determined via inductively coupled plasma mass spectrometry in 867 adults from Shiyan, China. Multivariable logistic/linear regression, restricted cubic spline regression, weighted quantile sum (WQS) regression and Bayesian kernel machine regression (BKMR) analyses were employed for exploring associations of individual metals or metal mixtures with hypertension prevalence and blood pressures. In total, 239 participants (29.5

Medical nutritional therapy, an evidence-based application of the nutrition care process guided by a registered dietitian nutritionist, is an integral part of diabetes management, with emerging evidence suggesting distinct nutritional requirements in this population. A thorough review of the existing literature reveals an imperative to address the specific requirements for vitamins and minerals in this population. Current guidelines have given limited attention to micronutrients, despite the higher prevalence of deficiencies and the altered patterns of association between serum micronutrient concentrations and multiple health outcomes among individuals with type 2 diabetes (T2D). To enhance diabetes management, an international multidisciplinary panel of 20 experts from 12 countries participated in a modified Delphi process, which was informed by a narrative review conducted by the research team, to reach consensus on 18 statements, encompassing 11 nutrient-specific recommendations and 7 statements outlining future directions in diabetes nutrition therapy research. The plan for this consensus has been registered on the Practice Guideline Registration for Transparency (registration number: PREPARE-2025CN1178). The expert panel proposed potential target serum concentrations, screening strategies, and micronutrient supplementation for people with T2D and that personalized nutritional strategies integrating individual characteristics, genetic information, and gut microbiota represent key areas for future research.

OBJECTIVES:Its uncertain how rapidly the detrimental effects of smoking on depression decline after cessation. Hence, we investigated the association between changes in smoking habits and late-life depressive symptoms. METHODS:We used data from 5247 men in the Singapore Chinese Health Study. Smoking was assessed at baseline (mean age of 52.5 years; 1993-98), follow-up 1 (mean age of 58.7 years; 1999-2004), follow-up 2 (mean age of 65.5 years; 2006-10) and follow-up 3 interviews (mean age of 72.8 years; 2014-16). Participants were categorized according to smoking status at baseline and subsequent follow-ups. Late-life depressive symptoms were measured using the 15-item Geriatric Depression Scale at follow-up 3. Logistic regression models were used to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) of depressive symptoms. RESULTS:Compared to never-smokers, while the multivariable-adjusted OR (95% CI) for depression was 2.19 (1.78-2.70) in persistent smokers, this risk was reduced progressively with increasing duration of cessation. Specifically, the OR (95% CI) was 1.39 (1.03-1.88) in those quitting for <10 years and substantially reduced to 1.11 (0.86-1.44) in those quitting for ≥30 years. Expectantly, earlier age at cessation was associated with a greater reduction in the likelihood; the OR (95% CI) was 1.39 (1.08-1.79) in those quitting 60 years and older, whereas it was 1.06 (0.79-1.43) in those quitting before 40 years old compared to never-smokers. CONCLUSIONS:The deleterious effects of smoking on late-life depression were substantially reduced only after sustained quitting for ≥30 years or quitting before 40 years old.

CONTEXT:The association between copper exposure and gestational diabetes mellitus (GDM) remains inconclusive. OBJECTIVE:Our study aimed to investigate the prospective relationship between urinary copper and GDM and the mediating role of DNA methylation and proteomic biomarkers in this association. METHODS:A nested case-control study was conducted based on the Tongji-Huaxi-Shuangliu Birth Cohort. Urinary copper levels and genome-wide DNA methylation were measured in early pregnancy for 432 pregnant women, and 737 circulating proteins were measured in a subset of 150 pregnant women. RESULTS:Urinary copper levels were positively associated with risk of GDM (adjusted odds ratio = 1.48 for each 1-unit increase in the log-transformed levels of copper, 95% CI, 1.15-1.91). A total of 73 differential cytosine-phosphoguanine sites (CpGs) were identified as associated with copper. Of these CpGs, cg23773809 annotated to SNX10 mediated 24.7% and 22.4% of the copper-GDM and copper-1 hour plasma glucose (1-h PG) association, respectively. cg04168577 annotated to PPFIBP2 and cg06105935 located in the intergenic region mediated 23.4% and 13.9% of the copper-1-h PG association, respectively. The protein CD2AP was found to be a reliable predictor for GDM. The 73 differential CpGs mediated 64.1% of the copper-CD2AP association. CONCLUSION:Copper exposure may induce alterations in DNA methylation patterns, which can subsequently lead to changes in the expression of proteins associated with GDM and elevate the risk of developing GDM.

Temporal alterations in the host lipidome and gut microbiome, and their dynamic interactions, in the progression of antenatal depression remain poorly understood. Here, we characterized the maternal lipidome and gut microbiome in 1086 Chinese pregnant women from the Tongji-Huaxi-Shuangliu Birth Cohort using 2954 serum and 2812 fecal samples collected longitudinally across early, mid, and late pregnancy. Our analyses identified 95 host lipids, and seven gut microbial genera associated with antenatal depression. Period-specific and longitudinal analyses indicated that lysophosphatidylethanolamines (LPEs) exhibited concordant associations with antenatal depression. In addition, temporal analyses suggested that decreases in the abundances of gut microbial genera [Ruminococcus]_gnavus_group and Enterococcus preceded the onset of antenatal depression and demonstrated extensive associations with LPEs and acylcarnitines (CARs). Integrative analyses revealed bidirectional concurrent and temporal crosstalk between microbiota-lipid axis and antenatal depression, with LPEs and medium- and long-chain CARs constituting key molecules of this axis. Collectively, this study offers valuable insights into the interplay between host lipidome and gut microbiome in the development of antenatal depression, laying a foundation for future investigations into the microbiota-lipid axis during pregnancy.

Background: Adverse social determinants of health (SDoH) are associated with a higher risk of mortality; however, the biological mechanisms underlying this association remain poorly understood. We sought to compare associations of 11 epigenetic age acceleration metrics (EAAs) with SDoH and examine whether EAAs mediate association between SDoH and all cause and cardiovascular disease (CVD) mortality. Methods: We included 1757 participants (aged ≥50 years) who participated in the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2002. The mortality cases were identified through linkage to the National Death Index through December 31, 2019. Self-reported SDoHs across five domains were collected in each survey cycle. Each SDoH item was scored on a scale from 0 (favorable level) to 1 (unfavorable level). Total combined SDoH scores were calculated as the sum of all items, with higher scores indicating greater disadvantage. Eleven EAAs covering three generations of epigenetic clocks were calculated. Cox models were used to evaluate associations of combined SDoH and EAAs with mortality risk. Mediation analysis was performed to assess whether and to what extent EAAs mediated the association between combined SDoH and mortality. We also estimated the life expectancy according to status of combined SDoH and accelerated epigenetic aging. Findings: Higher combined SDoH scores (indicating greater disadvantage) were associated with faster EAAs primarily in second- and third-generation clocks, but not in most first-generation clocks. Each 1-point increment in the combined SDoH score was associated with higher PhenoAge acceleration (β, 0.06; 95% CI, 0.02-0.09), GrimAgeMort acceleration (β, 0.10; 95% CI, 0.07-0.14), GrimAge2Mort acceleration (β, 0.11; 95% CI, 0.07-0.15), and DunedinPoAm (β, 0.08; 95% CI, 0.04-0.13). Unfavorable SDoH was associated with a higher risk of all-cause (hazard ratio 1.40 [95% CI 1.13-1.74]) and CVD mortality (hazard ratio 1.55 [95% CI 1.09-2.21]), corresponding to a 3.2-year life expectancy reduction at age 50. EAAs significantly mediated the associations between combined SDoH and mortality, with the largest mediated proportions observed for GrimAgeMort and GrimAge2Mort acceleration, accounting for 27.9% and 29.3% of SDoH-associated all-cause mortality risk, respectively. Furthermore, compared with participants with favorable SDoH and non-accelerated epigenetic aging, those with both unfavorable SDoH and accelerated epigenetic aging had a 4.2- to 7.1-year shorter life expectancy at age 50 years. Interpretation: Our findings suggest that EAAs may provide the mechanistic link between adverse SDoH and higher risk of mortality.

Glucosamine is a commonly used supplement, but concerns persist regarding its use in individuals with diabetes. This study aimed to examine the association of habitual glucosamine use with risks of all-cause and cause-specific mortality among individuals with type 2 diabetes, and to explore the potential mediating role of blood biomarkers in these associations. We included 19125 participants with type 2 diabetes from the UK Biobank who provided self-reported information on glucosamine use at baseline. Cox proportional hazards models were used to estimate the associations, with results presented as hazard ratios (HRs) and its 95% confidence intervals (CIs). Mediation analyses were conducted to estimate the extent to which selected blood biomarkers may be involved in these associations. Over a median follow-up of 13.5 years, 3045 deaths were recorded, including 1095 cancer deaths and 741 cardiovascular deaths. In multivariable-adjusted analyses, glucosamine use was associated with lower risks of all-cause, cancer, and cardiovascular mortality, with HRs (95% CIs) of 0.77 (0.68, 0.86), 0.82 (0.68, 0.99), and 0.73 (0.57, 0.94), respectively. These associations were generally consistent across several subgroups, including diabetes-related factors, demographic characteristics, lifestyle, medication use and health conditions. Furthermore, blood biomarkers (such as cystatin C, alkaline phosphatase, and C-reactive protein) explained 13.61% (95% CI, 8.26, 34.14) of the association between glucosamine use and all-cause mortality. Our findings suggest that glucosamine use was inversely associated with mortality in individuals with type 2 diabetes, with several blood biomarkers partly explaining this association; however, clinical trials are needed to confirm these findings.

Background::Hypertensive disorders of pregnancy (HDP) are leading causes of maternal and fetal morbidity and mortality. Although lower grip strength has been linked to higher cardiovascular risk in the general population, evidence on its relationship with HDP remains limited. We aimed to prospectively investigate this association in a large cohort of pregnant women.Methods::Between March 2017 and January 2020, 6802 pregnant women (mean age ± standard deviation: 26.6 ± 3.7 years) enrolled in the Tongji-Huaxi-Shuangliu Birth Cohort were included in this analysis. Grip strength was measured in early pregnancy (mean gestational week ± standard deviation: 10.3 ± 2.0 weeks) and assessed in three ways: absolute grip strength (AGS) and two relative indices (AGS normalized to body mass index or body weight). Logistic regression models were used to assess the associations between grip strength and the risk of HDP (gestational hypertension or preeclampsia). Multiple blood biomarkers (blood lipids, leptin, adiponectin, C-reactive protein, C-peptide, glycated hemoglobin, and homeostatic model assessment of insulin resistance) were measured among 638 women at enrollment.Results::A total of 180 women developed HDP during pregnancy. The adjusted odds ratios (95% confidence intervals) of HDP across increasing quartiles of AGS in early pregnancy were 1.00 (reference), 0.93 (0.63-1.35), 0.67 (0.44-1.00), and 0.35 (0.21-0.56). Odds ratios across quartiles of two relative grip strength (RGS) indices showed similar patterns. Restricted cubic spline analyses indicated a nonlinear association between AGS and HDP risk ( P for nonlinearity = 0.014). The risk plateaued at lower AGS levels but showed a linear inverse association above a threshold of 18.1 kg. In contrast, the two relative measures showed linear associations with HDP risk ( P for nonlinearity ≥0.164). Additionally, higher RGS was generally correlated with favorable metabolic profiles (e.g., lower levels of low-density lipoprotein cholesterol, triglycerides, and C-reactive protein). Conclusion::Grip strength in early pregnancy was inversely associated with the risk of HDP, and RGS may serve as a simple and useful measure for risk stratification.

Objective:To examine the association between overall socioeconomic status (SES) and incident diabetes, to estimate how much of the SES-diabetes association is explained by modifiable diabetes risk factors, and to assess whether the benefits of favorable risk factor profiles differ by SES. Methods:We analyzed 337,229 adults without diabetes at baseline from the UK Biobank. Overall SES was derived using latent class analysis based on income, occupation, and education. Modifiable diabetes risk factor scores were constructed across physiological, behavioral, environmental, and psychological domains. Cox proportional hazard models and additive hazard models were used to evaluate associations, mediation proportions, and interactions for incident diabetes. Results:During a median follow-up of 12.5 years, 11,557 participants developed diabetes ascertained through linkage to registries. The low SES group had 2.47-fold (95% CI: 2.33-2.62) diabetes risk and 2.7 (2.5-2.8) more incident diabetes cases per 1,000 person-years compared to the high SES group, 54.4% of which was explained by all modifiable factors jointly, with physiological score contributing to the largest proportion (39.1%). Favorable risk factor profiles were associated with lower diabetes risk across all SES groups, and absolute risk reductions associated with favorable profiles were greatest among individuals with low SES ( P for additive interaction ≤ 0.002). Conclusion:More than half of the excess diabetes risk associated with low SES can be explained by modifiable risk factors. Improving these factors may contribute to greater reduction in diabetes incidence among socioeconomically disadvantaged populations, supporting targeted diabetes prevention strategies to reduce socioeconomic disparities.

Preterm birth (PTB) remains a major pregnancy complication, yet the role of the maternal gut virome in its etiology is largely unknown. Here we show that the maternal gut virome undergoes ecological destabilization prior to PTB, coupled with distinct host metabolic remodeling. Nested within the Tongji-Huaxi-Shuangliu Birth Cohort, we integrate longitudinal gut virome and bacteriome profiles from 300 stool samples, alongside matched serum metabolomes and clinical profiles, from 100 pregnant women (50 with PTB and 50 with term birth) across early, middle, and late pregnancy. We reveal that although the maternal gut virome is highly personalized and longitudinally stable within individuals, PTB is characterized by reduced virome convergence and specific alterations in viral populations emerging during mid-to-late pregnancy. Host-phage analyses identify remodeling of Klebsiella- and Prevotella-associated viral communities linked to PTB risk. PTB-associated virome alterations are further associated with amino acid metabolic remodeling, particularly glutamate- and aspartate-related pathways, supported by reproducible virus-metabolite associations and enriched viral auxiliary metabolic genes. In addition, L-aspartate partly mediates associations between monocyte-related inflammatory indices and PTB. Multi-omics modeling demonstrates that virome-metabolome signatures achieve strong predictive performance for both PTB and imminent delivery, with viral features contributing substantially to prediction accuracy and retaining predictive value in external validation. Collectively, these findings highlight the maternal virome and its metabolic signatures as key determinants of PTB susceptibility.