GRAPHICAL ABSTRACT

Highlights

• Fat distribution mediates the association between birth weight and T2DM.

• Liver fat is the strongest mediator, independent of overall adiposity.

• Low birth weight is linked to greater visceral and ectopic fat in adulthood.

• Fat distribution–diabetes pathways show clear sex-specific patterns.

• Findings support early risk profiling based on fat topography rather than BMI alone.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) presents a growing global health challenge, with its incidence rising rapidly across both developed and developing countries [1]. While obesity is a well-established risk factor, emerging evidence suggests that the distribution of fat—rather than total fat mass—is a more robust predictor of T2DM risk [2,3]. In particular, excess visceral adipose tissue (VAT), located deep within the abdominal cavity, is strongly linked to insulin resistance and T2DM, independent of body mass index (BMI) [4]. Individuals with similar BMIs may display markedly different metabolic profiles depending on fat distribution, underscoring the need to consider regional fat depots in diabetes risk assessment [5,6].

Although fat distribution is influenced by genetic and lifestyle factors, increasing attention has been directed toward early-life exposures as critical determinants of long-term metabolic health [7]. Intrauterine conditions, which shape fetal growth and organ development, are thought to influence adipose tissue development and function across the lifespan [8]. Birth weight (BW), a proxy for intrauterine growth, reflects the net impact of fetal nutrition, maternal health, and prenatal environment [9]. Previous studies have reported consistent associations between low BW (<2.5 kg) and an elevated T2DM risk in adulthood [10,11]. Findings from famine studies further support this association, showing that prenatal exposure to severe nutritional deficiency was associated with an increased risk of T2DM later in life [12,13]. In contrast, the relationship between high BW (>4.0 kg) and T2DM risk is less consistent, with studies reporting both increased [11] and decreased risks [10].

Past investigations have linked BW to crude measures of fat distribution such as waist circumference or waist-to-hip ratio [14]. While informative, these indices cannot differentiate between specific fat depots or control for the confounding of overall adiposity. Importantly, fat distribution exhibits clear sexual dimorphism: women typically accumulate more gluteofemoral and subcutaneous fat, whereas men are more prone to visceral and ectopic fat storage [15,16]. It remains unclear whether early-life factors like BW alter these sex-specific fat distribution patterns and whether such changes mediate the risk of T2DM.

In this study, we aimed to (1) assess the associations between BW, adult fat distribution, and T2DM risk; (2) evaluate whether fat depots mediate the BW-T2DM association; and (3) examine sex-specific differences in these associations. Leveraging high-resolution magnetic resonance imaging (MRI) and dual-energy X-ray absorptiometry (DXA) imaging data from the UK Biobank (UKB), we precisely quantified regional fat depots—including visceral, subcutaneous, gynoid, hepatic, and pancreatic fat, and muscle fat infiltration (MFI)—and adjusted each for BMI to isolate distributional effects. Our study contributes to the understanding of developmental origins of T2DM and highlights the potential of fat distribution profiling to identify high-risk individuals before overt metabolic dysfunction occurs.

METHODS

Study population

This study was conducted using data from the UKB, a large prospective cohort that enrolled over 500,000 participants aged 40 to 69 years across the UK between 2006 and 2010 [17]. For the current analysis, we included participants who underwent either MRI or DXA imaging between April 2014 and September 2019 [18]. We excluded individuals with missing BW data (n=222,197), missing fat depot measurements (n=248,669), or a diagnosis of diabetes at baseline (n=786). The final analytic sample consisted of 30,718 diabetes-free participants.

Exposure assessment

BW was self-reported at baseline in either kilograms or imperial units and was converted into kilograms for consistency. The validity of self-reported BW in UKB has been previously demonstrated [19]. In our study, participants from both singleton and multiple pregnancies were included. Approximately 2.3% of participants in our analytic sample reported being from multiple pregnancies. Since our aim was to evaluate the overall association between BW and later T2DM risk, we did not distinguish the underlying causes of low BW (e.g., intrauterine growth restriction vs. multiple gestation).

Fat depot measurements

Regional fat depots were assessed using high-resolution imaging modalities. VAT and abdominal subcutaneous adipose tissue (ASAT) were quantified from MRI scans acquired using the Siemens Aera 1.5-T scanner and analyzed with the AMRA Profiler Research software (AMRA Medical, Linköping, Sweden). VAT was defined as adipose tissue within the abdominal cavity, excluding adipose located outside the skeletal muscles and posterior to the spine. ASAT was defined as subcutaneous adipose tissue in the abdomen from the top of the femoral head to vertebral level T9. Fat volumes were converted to mass using a specific density of 0.9 g/mL to facilitate interpretation [20].

Liver and pancreatic fat were quantified using proton density fat fraction maps derived from single-slice multi-echo Dixon MRI sequences. Representative regions of interest were selected to calculate liver fat fraction (LFF) and pancreatic fat fraction (PFF) [21,22]. MFI of the anterior thigh was quantified as the fat fraction within viable muscle tissue. Additional imaging details have been described previously [23].

Gynoid adipose tissue (GAT) was assessed via whole-body DXA scans (Lunar iDXA, GE Healthcare, Chicago, IL, USA) following standard positioning protocols [24]. Detailed documentation on the DXA procedures is available online at UKB DXA documentation.

Covariates

Baseline covariates included age, sex, ethnicity, assessment center, household income, education level, sleep duration, smoking status, alcohol use, physical activity (measured in metabolic equivalent of task-minutes/week), family history of diabetes, diet score, index of multiple deprivation (IMD), and BMI. The diet score, reflecting adherence to a healthy dietary pattern [25], was derived by assigning points for favorable consumption of key food groups (e.g., fruits, vegetables, whole grains, fish) and deducting points for unfavorable items (e.g., processed meats, unprocessed red meat, refined grains) [26]. Higher scores indicate healthier dietary habits. IMD is a composite measure of socioeconomic status used in the UKB to quantify area-level deprivation. It is derived from national administrative data and combines information across several domains, including income, employment, education, health, crime, barriers to housing and services, and the living environment. Each participant in the UKB is assigned an IMD score based on their residential postcode at baseline, with higher scores indicating greater deprivation. Sedentary behaviors were assessed by self-reported sitting time (including driving, computer use, and TV watching). The proportions of missing data for covariates were no more than 2%. Median and mode imputation were used for continuous and categorical covariates, respectively.

Outcome ascertainment

Incident T2DM was ascertained via linkage to hospital inpatient records, primary care data, death registries, and self-reported diagnoses, coded according to International Classification of Diseases (ICD)-9 and ICD-10. The earliest record across these sources was taken as the date of incident diabetes. The follow-up duration for each participant was calculated from the date of enrollment until first occurrence of diabetes, death, loss to follow-up, or the end of the study period (December 19, 2022), whichever occurred first.

Statistical analysis

We followed a stepwise analytical strategy to investigate the associations among BW, regional fat distribution, and incident T2DM, as well as the mediating role of specific fat depots. To evaluate the association between BW and the risk of T2DM, we first treated BW as a continuous variable and modeled its relationship with T2DM risk using Cox proportional hazards regression with restricted cubic splines (knots at the 10th, 50th, and 90th percentiles) to assess nonlinearity. We then categorized BW into three groups: low birth weight (LBW, <2.5 kg), normal birth weight (NBW, 2.5–4.0 kg, reference), and high birth weight (HBW, >4.0 kg), in line with World Health Organization and obstetric guideline definitions of LBW and macrosomia [27–29]. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated for each group. The proportional hazards assumption was verified using Schoenfeld residuals.

Covariates were progressively adjusted in multivariable models. Model 1 adjusted for age and sex to account for basic demographic confounding. Model 2 further adjusted for ethnicity, assessment center, IMD, education, household income, smoking status, alcohol consumption, sleep duration, diet score, physical activity, sedentary behaviors, and family history of diabetes. These variables were included because they are established determinants of T2DM and could confound the relationship between BW and later-life outcomes. Model 3 additionally adjusted for BMI to examine whether the associations were independent of overall adiposity.

Fat depots—particularly visceral and ectopic fat—are strongly correlated with BMI, which may confound their associations with metabolic outcomes. To examine potential multicollinearity between BMI and fat depots, we calculated variance inflation factors (VIFs) for BMI and each fat depot measure. VIF values ranged from 2.02 to 2.46, indicating no significant multicollinearity. Furthermore, individuals with similar BMI levels can exhibit distinct fat distribution patterns and metabolic risks. To isolate the role of fat distribution independent of overall adiposity, we adjusted each fat depot for BMI using sex-specific linear regression models. The resulting residuals (i.e., BMI-adjusted fat depots) represent variation in fat storage that is not explained by BMI. These residuals were standardized to generate standard deviation (SD) scores for use in downstream analyses and mediation modeling.

We assessed whether BW was associated with adult fat distribution patterns by modeling the relationship between categorical BW and each BMI-adjusted fat depot, including visceral adipose tissue (VATadjBMI), abdominal subcutaneous fat tissue (ASATadjBMI), gynoid fat tissue (GATadjBMI), liver fat fraction (LFFadjBMI), pancreatic fat fraction (PFFadjBMI), and muscle fat infiltration (MFIadjBMI). Linear regression models were used to estimate the beta coefficients and 95% CIs. Given the right-skewed distributions of LFF, PFF, and MFI and evidence from residual diagnostics of the initial linear models indicating violations of normality and homoscedasticity, we applied a natural log transformation to these variables prior to regression analyses. To examine the association between each BMI-adjusted fat depot and incident T2DM, we applied Cox proportional hazards models. Analyses were adjusted for covariates as defined in model 1–3.

To examine whether regional fat distribution mediates the association between BW and T2DM, we conducted formal mediation analyses using a regression-based approach proposed by Li et al. [30] and VanderWeele [31]. For each fat depot, we constructed two models: a logistic regression model predicting T2DM from BW and the fat depot (mediator) and a linear regression model predicting the fat depot from BW. Both models were adjusted for the covariates included in model 3. The indirect effect (mediated by fat depots) and direct effect (not mediated) were estimated, and the proportion mediated was calculated on the odds ratio scale using 1,000 bootstrap resamples to obtain bias-corrected CIs.

In the UKB, fat depots were measured several years after baseline enrollment. To address potential bias from this time lag, we conducted sensitivity analyses excluding participants who developed T2DM before the imaging visit and cases diagnosed within 2 years after imaging. These restrictions reduced the possibility that preclinical or early-onset T2DM influenced fat distribution at the time of imaging.

All analyses were conducted in R version 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria), using two-sided P values and 95% CIs for statistical inference.

Ethics approval and consent to participate

The data analyzed in this study were obtained from the UKB, which was approved by the North West Multi-center Research Ethics Committee (REC reference: 21/NW/0157). All participants provided written informed consent.

Availability of data and material

This research was conducted using the UKB Resource under Application Number 98973. Researchers can request access to the UKB data by submitting an application via the UKB website (https://www.ukbiobank.ac.uk).

RESULTS

Participant characteristics

A total of 30,718 diabetes-free individuals from the UKB were included in the final analysis. A total of 771 participants developed T2DM during a median follow-up of 13.9 years. The mean age at baseline was 53.9 years, and 41.6% were men. BW distribution was as follows: 8.2% had LBW, 78.7% had NBW, and 13.1% had HBW. Participants with LBW were more likely to be women, have lower income, and higher deprivation scores, while those with HBW tended to have higher BMI and were more often men (Table 1).

Association between birth weight and risk of T2DM

We first modeled BW as a continuous variable to explore its relationship with incident T2DM. A linear relationship was observed between BW and T2DM risk, with each 1 kg increase in BW associated with a 29% lower risk of T2DM (HR, 0.71; 95% CI, 0.64 to 0.79) (Fig. 1A). Sex-stratified analyses revealed notable differences. In women, the inverse association was linear and stronger, while in men, the relationship was weaker and followed an L-shaped pattern (Fig. 1B and C). When BW was categorized into low, normal, and high groups, LBW was associated with a significantly higher T2DM risk compared to NBW (HR for women, 1.76 [95% CI, 1.32 to 2.35]; HR for men, 1.63 [95% CI, 1.19 to 2.23]; P for interaction=0.56), whereas HBW was associated with a lower risk (HR for women, 0.52 [95% CI, 0.33 to 0.81]; HR for men, 0.77 [95% CI, 0.59 to 0.99], P for interaction=0.09) (Table 2). Analyses treating BW as a continuous variable yielded significant sex differences (P for interaction=0.03) (Table 2).

Association between birth weight and BMI-adjusted fat depots

We then assessed whether BW was associated with fat distribution in adulthood by modeling BW as a categorical variable in relation to BMI-adjusted fat depots. Significant correlations were observed between fat depots and BMI in both women and men (Fig. 2A). Thus, it is necessary to distinguish the unique impact of each fat depot from overall BMI. In linear regression models, LBW was associated with higher levels of VATadjBMI, ASATadjBMI, LFFadjBMI, PFFadjBMI, and lower levels of GATadjBMI, compared to NBW. Conversely, HBW was associated with lower VATadjBMI, LFFadjBMI, and PFFadjBMI, but higher GATadjBMI (Fig. 2B). Sex-specific analyses revealed that the associations between BW and fat distribution were more pronounced in women than in men (Fig. 2C and D, Supplementary Table 1).

Association between BMI-adjusted fat depots and T2DM risk

We next investigated whether BMI-adjusted fat depots were independently associated with incident T2DM. Higher levels of VATadjBMI (HR per 1 SD increase, 1.36; 95% CI, 1.26 to 1.47), LFFadjBMI (HR, 1.34; 95% CI, 1.27 to 1.41), PFFadjBMI (HR, 1.19; 95% CI, 1.10 to 1.28), and MFIadjBMI (HR, 1.11; 95% CI, 1.03 to 1.19) were significantly associated with an increased risk of T2DM. In contrast, GATadjBMI was inversely associated with T2DM risk (HR, 0.83; 95% CI, 0.77 to 0.90), while ASATadjBMI showed neutral association (HR, 0.95; 95% CI, 0.88 to 1.03). Notably, the association between VATadjBMI and T2DM risk was stronger in women (HR, 1.86; 95% CI, 1.61 to 2.15) than in men (HR, 1.21; 95% CI, 1.11 to 1.33; P for interaction <0.001) (Table 3). Sensitivity analyses excluding participants who developed diabetes prior to imaging or were diagnosed within 2 years thereafter yielded similar results, supporting the robustness of the findings (Supplementary Table 2).

Mediating role of fat depots in the relationship between birth weight and T2DM risk

We conducted mediation analysis to determine whether regional fat depots mediated the association between BW and T2DM. Liver fat (LFFadjBMI) exhibited the strongest mediating effect, accounting for 11% of the association between each 1 SD decrease in BW and increased T2DM risk, followed by VATadjBMI (5%), GATadjBMI (4%), PFFadjBMI (3%), and MFIadjBMI (1%). In sex-stratified analyses, the BW-T2DM association in women was primarily mediated by LFFadjBMI (12%), VATadjBMI (7%), PFFadjBMI (3%), and GATadjBMI (2%). In men, the main mediators were LFFadjBMI (9%), GATadjBMI (8%), PFFadjBMI (3%), and VATadjBMI (3%) (Table 4).

DISCUSSION

In this large population-based study of over 30,000 diabetes-free adults, we found that adult fat distribution—specifically increased visceral, liver, and pancreatic fat and reduced gluteofemoral fat—plays a key mediating role in the association between BW and incident T2DM. These associations were independent of total adiposity, as all fat depots were adjusted for BMI. Among the fat depots examined, liver fat showed the strongest mediating effect, accounting for 11% of the BW-diabetes association. Notably, the patterns of mediation differed by sex: visceral and liver fat were predominant mediators in women, while liver and gluteofemoral fat contributed more in men. These findings highlight the importance of fat distribution—rather than overall fat mass—as a mechanistic link between early-life growth and adult metabolic disease.

Birth weight and risk of T2DM

Our findings align with previous studies showing that LBW is associated with a significantly higher risk of developing T2DM later in life [10,11,32], consistent with the Barker hypothesis that intrauterine growth restriction programs long-term metabolic health [33]. By contrast, the relationship between HBW and T2DM risk is less well understood and remains controversial [10,11]. While some studies suggest a J-shaped relationship, with both low and high extremes of BW linked to elevated risk [11], in our study HBW was associated with a reduced risk of T2DM. One possible explanation is that HBW participants in UKB may represent a relatively healthier subgroup, reflecting favorable fetal growth rather than maternal diabetes or fetal overnutrition, both of which are established risk factors for T2DM. This interpretation, however, cannot be directly confirmed due to limited information on early-life exposures. Furthermore, potential misclassification of self-reported BW decades after birth and the smaller proportion of HBW participants may have attenuated associations.

We also observed a sex-specific pattern: the BW-T2DM association was stronger in women than in men. This finding aligns with prior research suggesting that females may be more susceptible to early-life environmental influences. Potential mechanisms include sex differences in fetal growth trajectories, placental function, and epigenetic programming [34,35].

Intrauterine development and adult fat distribution

Previous studies have shown that individuals with LBW tend to accumulate more visceral fat, less GAT, and have a lower muscle-to-fat ratio [36–38], indicating that intrauterine development is critical determinant of adult fat distribution [39]. Our findings are consistent with these observations and further highlight important sex differences. Women generally exhibit greater subcutaneous and gluteofemoral fat and less visceral and ectopic fat than men, which may underlie their differential susceptibility to the metabolic consequences of LBW. In contrast, the associations between LBW and fat distribution in men did not reach statistical significance, likely due to the smaller number of LBW men, which limited statistical power.

Fat distribution and T2DM risk

BMI is a widely used measure of adiposity but does not capture regional fat distribution [5]. By adjusting for BMI, we were able to isolate the independent effects of specific fat depots on T2DM risk [4]. Our results emphasize that it is not merely the amount of fat but its anatomical distribution that shapes metabolic outcomes. This highlights the importance of considering fat distribution patterns, rather than total fat mass alone, in understanding the pathophysiology of T2DM.

Specifically, we found that VAT, LFF, and PFF were strongly associated with increased T2DM risk, likely due to their secretion of pro–inflammatory cytokines and adverse adipokines that promote systemic inflammation, insulin resistance and metabolic dysfunction–processes that plausibly mediate the BW-T2DM association [40]. In contrast, GAT appeared protective, which may reflect its lower lipolytic activity, reduced secretion of inflammatory adipokines, greater capacity for long-term fatty acid storage, and release of beneficial adipokines such as leptin and adiponectin [40].

Mediating role of fat distribution in the birth weight–diabetes relationship

A key contribution of our study is the demonstration that adult fat distribution partially mediates the association between BW and T2DM. Visceral and ectopic fat depots, particularly in liver and pancreas, accounted for a significant proportion of the excess T2DM risk linked to LBW. This supports the hypothesis that early-life growth patterns influence not only total fat accumulation but also its distribution, which in turn shapes long-term metabolic risk.

Moreover, fat distribution exhibits pronounced sexual dimorphism. In our analysis, the mediating effects of LFF and VAT on the association between BW and T2DM were more evident in women (12% and 7%, respectively) than in men (9% and 3%), highlighting sex-specific pathways through which LBW contributes to metabolic vulnerability. These findings suggest that incorporating fat distribution into risk stratification may improve the identification of individuals at higher metabolic risk following adverse early-life conditions such as LBW, thereby informing more effective prevention strategies.

Strengths and limitations

Strengths of this study include the large sample size, high-resolution imaging of fat depots, and comprehensive adjustment for potential confounders. The ability to adjust for BMI and isolate the independent effects of fat distribution on T2DM risk represents a significant advancement over previous studies that have primarily focused on total fat mass.

Several limitations should be acknowledged. First, the UKB cohort is predominantly of European ancestry, which limits the generalizability of our findings to other populations. Second, BW was self-reported decades later, introducing potential recall bias and misclassification, which might have attenuated the BW-T2DM associations. Third, residual confounding by unmeasured factors cannot be excluded, despite adjustment for potential confounders. Finally, adult socioeconomic and lifestyle factors, as well as BMI, may act as mediators rather than mere confounders, and thus our estimates of the BW-T2DM association could be conservative due to potential over-adjustment.

Conclusions

Our findings demonstrate that lower BW is associated with adverse fat distribution in adulthood—particularly excess liver and visceral fat—which subsequently increases the risk of T2DM. These associations were more pronounced in women than in men. Accordingly, individuals, especially women, with documented low BW may warrant closer metabolic surveillance, including imaging-based assessment of fat distribution, to complement conventional BMI-based risk stratification.

SUPPLEMENTARY MATERIALS

NOTES

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

AUTHOR CONTRIBUTIONS

Conception or design: X.Y., J.L., R.A.

Acquisition, analysis, or interpretation of data: D.D., X.L., S.C., Z.L.

Drafting the work or revising: D.D., X.K., J.L.

Final approval of the manuscript: all authors.

FUNDING

Ding Ding was supported by the National Natural Science Foundation of China (NSFC) (82304116). Jie Li was supported by NSFC (82473620), National Science and Technology Major Project (2025ZD0549500), National High-Level Talent Special Support Program for Young Top-Notch Talents, Start-up Funding of Guangdong Provincial People’s Hospital, and Guangzhou Science and Technology Program (2025A03J4433). Xingfen Yang was supported by NSFC (82373600). The funding sources do not have any role in the design, interpretation of the study, or the decision to publish the results.

ACKNOWLEDGMENTS

The authors appreciate all UKB participants and all staff for their contribution to these studies.

Fig. 1

Nonlinear relationship between birth weight and type 2 diabetes mellitus risk among 30,718 UK Biobank participants. Birth weight was treated as a continuous variable (per kg) and modeled its relationship with type 2 diabetes mellitus risk using Cox proportional hazards regression with restricted cubic splines (knots at the 10th, 50th, and 90th percentiles) to assess nonlinearity, with adjustment the same as model 3 in total population (A), women (B) and men (C). HR, hazard ratio; CI, confidence interval.

Fig. 2

Correlation plots of local fat depots and body mass index (BMI)-adjusted fat depot comparison between birth weight groups. Part figure (A) displays the correlation plots between local fat depots and BMI in total population. Part figures (B–D) respectively display sex-specific BMI-adjusted fat depot comparison between birth weight groups in total population (B), women (C) and men (D), based on linear regression models adjusted for covariates in model 3. Liver fat, pancreatic fat, and muscle fat infiltration were natural log transformed prior to regression analyses. Birth weight was categorized as low (<2.5 kg), normal (2.5–4.0 kg), and high (>4.0 kg). VAT, visceral adipose tissue; ASAT, abdominal subcutaneous adipose tissue; GAT, gynoid adipose tissue; LFF, liver fat fraction; PFF, pancreatic fat fraction; MFI, muscle fat infiltration; BMI, body mass index; VATadjBMI, visceral adipose tissue adjusted for BMI; ASATadjBMI, abdominal subcutaneous adipose tissue adjusted for BMI; GATadjBMI, gynoid adipose tissue adjusted for BMI; LFFadjBMI, liver fat fraction adjusted for BMI; PFFadjBMI, pancreatic fat fraction adjusted for BMI; MFIadjBMI, muscle fat infiltration adjusted for BMI; LBW, low birth weight; NBW, normal birth weight; HBW, high birth weight.

REFERENCES

• 1. GBD 2021 Diabetes Collaborators. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2023;402:203-34.PubMedPMC
• 2. Yamazaki H, Tauchi S, Machann J, Haueise T, Yamamoto Y, Dohke M, et al. Fat distribution patterns and future type 2 diabetes. Diabetes 2022;71:1937-45.ArticlePubMedPDF
• 3. Neeland IJ, Ross R, Despres JP, Matsuzawa Y, Yamashita S, Shai I, et al. Visceral and ectopic fat, atherosclerosis, and cardiometabolic disease: a position statement. Lancet Diabetes Endocrinol 2019;7:715-25.ArticlePubMed
• 4. Agrawal S, Klarqvist MD, Diamant N, Stanley TL, Ellinor PT, Mehta NN, et al. BMI-adjusted adipose tissue volumes exhibit depot-specific and divergent associations with cardiometabolic diseases. Nat Commun 2023;14:266.ArticlePubMedPMCPDF
• 5. Stefan N. Causes, consequences, and treatment of metabolically unhealthy fat distribution. Lancet Diabetes Endocrinol 2020;8:616-27.ArticlePubMed
• 6. Song Z, Gao M, Lv J, Yu C, Guo Y, Bian Z, et al. Metabolically healthy obesity, transition to unhealthy phenotypes, and type 2 diabetes in 0.5 million Chinese adults: the China Kadoorie Biobank. Eur J Endocrinol 2022;186:233-44.ArticlePubMedPMC
• 7. Vaiserman A, Lushchak O. Developmental origins of type 2 diabetes: focus on epigenetics. Ageing Res Rev 2019;55:100957.ArticlePubMed
• 8. Ye J, Gao C, Liang Y, Hou Z, Shi Y, Wang Y. Characteristic and fate determination of adipose precursors during adipose tissue remodeling. Cell Regen 2023;12:13.ArticlePubMedPMCPDF
• 9. Hanson MA, Gluckman PD. Early developmental conditioning of later health and disease: physiology or pathophysiology? Physiol Rev 2014;94:1027-76.ArticlePubMedPMC
• 10. Wibaek R, Andersen GS, Linneberg A, Hansen T, Grarup N, Thuesen AC, et al. Low birthweight is associated with a higher incidence of type 2 diabetes over two decades independent of adult BMI and genetic predisposition. Diabetologia 2023;66:1669-79.ArticlePubMedPMCPDF
• 11. Knop MR, Geng TT, Gorny AW, Ding R, Li C, Ley SH, et al. Birth weight and risk of type 2 diabetes mellitus, cardiovascular disease, and hypertension in adults: a meta-analysis of 7 646 267 participants from 135 studies. J Am Heart Assoc 2018;7:e008870.ArticlePubMedPMC
• 12. Lumey LH, Li C, Khalangot M, Levchuk N, Wolowyna O. Fetal exposure to the Ukraine famine of 1932–1933 and adult type 2 diabetes mellitus. Science 2024;385:667-71.ArticlePubMed
• 13. Taeubert MJ, Kuipers TB, Zhou J, Li C, Wang S, Wang T, et al. Adults prenatally exposed to the Dutch Famine exhibit a metabolic signature associated with a broad spectrum of common diseases. BMC Med 2024;22:309.ArticlePubMedPMCPDF
• 14. Araujo de Franca GV, Restrepo-Mendez MC, Loret de Mola C, Victora CG. Size at birth and abdominal adiposity in adults: a systematic review and meta-analysis. Obes Rev 2014;15:77-91.ArticlePubMedPDF
• 15. Rask-Andersen M, Karlsson T, Ek WE, Johansson A. Genome-wide association study of body fat distribution identifies adiposity loci and sex-specific genetic effects. Nat Commun 2019;10:339.ArticlePubMedPMCPDF
• 16. Karpe F, Pinnick KE. Biology of upper-body and lower-body adipose tissue: link to whole-body phenotypes. Nat Rev Endocrinol 2015;11:90-100.ArticlePubMedPDF
• 17. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 2015;12:e1001779.ArticlePubMedPMC
• 18. Petersen SE, Matthews PM, Bamberg F, Bluemke DA, Francis JM, Friedrich MG, et al. Imaging in population science: cardiovascular magnetic resonance in 100,000 participants of UK Biobank: rationale, challenges and approaches. J Cardiovasc Magn Reson 2013;15:46.ArticlePubMedPMCPDF
• 19. Tyrrell JS, Yaghootkar H, Freathy RM, Hattersley AT, Frayling TM. Parental diabetes and birthweight in 236 030 individuals in the UK biobank study. Int J Epidemiol 2013;42:1714-23.ArticlePubMedPMC
• 20. Vogelezang S, Santos S, Toemen L, Oei EH, Felix JF, Jaddoe VW. Associations of fetal and infant weight change with general, visceral, and organ adiposity at school age. JAMA Netw Open 2019;2:e192843.ArticlePubMedPMC
• 21. Wilman HR, Kelly M, Garratt S, Matthews PM, Milanesi M, Herlihy A, et al. Characterisation of liver fat in the UK Biobank cohort. PLoS One 2017;12:e0172921.ArticlePubMedPMC
• 22. Triay Bagur A, Aljabar P, Ridgway GR, Brady M, Bulte DP. Pancreas MRI segmentation into head, body, and tail enables regional quantitative analysis of heterogeneous disease. J Magn Reson Imaging 2022;56:997-1008.PubMedPMC
• 23. Linge J, Borga M, West J, Tuthill T, Miller MR, Dumitriu A, et al. Body composition profiling in the UK biobank imaging study. Obesity (Silver Spring) 2018;26:1785-95.ArticlePubMedPMCPDF
• 24. Littlejohns TJ, Holliday J, Gibson LM, Garratt S, Oesingmann N, Alfaro-Almagro F, et al. The UK Biobank imaging enhancement of 100,000 participants: rationale, data collection, management and future directions. Nat Commun 2020;11:2624.ArticlePubMedPMCPDF
• 25. Mozaffarian D. Dietary and policy priorities for cardiovascular disease, diabetes, and obesity: a comprehensive review. Circulation 2016;133:187-225.ArticlePubMedPMC
• 26. Huang J, Ye E, Li X, Niu D, Wang J, Zhao Y, et al. Association of healthy diet score and adiposity with risk of colorectal cancer: findings from the UK Biobank prospective cohort study. Eur J Nutr 2024;63:2055-69.ArticlePubMedPDF
• 27. Ye J, Torloni MR, Ota E, Jayaratne K, Pileggi-Castro C, Ortiz-Panozo E, et al. Searching for the definition of macrosomia through an outcome-based approach in low- and middle-income countries: a secondary analysis of the WHO Global Survey in Africa, Asia and Latin America. BMC Pregnancy Childbirth 2015;15:324.ArticlePubMedPMCPDF
• 28. American College of Obstetricians and Gynecologists’ Committee on Practice Bulletins-Obstetrics. Practice Bulletin No. 173: fetal macrosomia. Obstet Gynecol 2016;128:e195-209.PubMed
• 29. Araujo E Junior, Peixoto AB, Zamarian AC, Elito J Junior, Tonni G. Macrosomia. Best Pract Res Clin Obstet Gynaecol 2017;38:83-96.ArticlePubMed
• 30. Li J, Glenn AJ, Yang Q, Ding D, Zheng L, Bao W, et al. Dietary protein sources, mediating biomarkers, and incidence of type 2 diabetes: findings from the women’s health initiative and the UK biobank. Diabetes Care 2022;45:1742-53.ArticlePubMedPMCPDF
• 31. VanderWeele TJ. Mediation analysis: a practitioner’s guide. Annu Rev Public Health 2016;37:17-32.ArticlePubMed
• 32. Olaiya MT, Wedekind LE, Hanson RL, Sinha M, Kobes S, Nelson RG, et al. Birthweight and early-onset type 2 diabetes in American Indians: differential effects in adolescents and young adults and additive effects of genotype, BMI and maternal diabetes. Diabetologia 2019;62:1628-37.ArticlePubMedPMCPDF
• 33. Hales CN, Barker DJ. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia 1992;35:595-601.ArticlePubMedPDF
• 34. Dearden L, Bouret SG, Ozanne SE. Sex and gender differences in developmental programming of metabolism. Mol Metab 2018;15:8-19.ArticlePubMedPMC
• 35. Braun AE, Mitchel OR, Gonzalez TL, Sun T, Flowers AE, Pisarska MD, et al. Sex at the interface: the origin and impact of sex differences in the developing human placenta. Biol Sex Differ 2022;13:50.ArticlePubMedPMCPDF
• 36. Ronn PF, Smith LS, Andersen GS, Carstensen B, Bjerregaard P, Jorgensen ME. Birth weight and risk of adiposity among adult Inuit in Greenland. PLoS One 2014;9:e115976.ArticlePubMedPMC
• 37. Honda M, Tsuboi A, Minato-Inokawa S, Takeuchi M, Yano M, Kurata M, et al. Birth weight was associated positively with gluteofemoral fat mass and inversely with 2-h postglucose insulin concentrations, a marker of insulin resistance, in young normal-weight Japanese women. Diabetol Int 2022;13:375-80.ArticlePubMedPDF
• 38. Kensara OA, Wootton SA, Phillips DI, Patel M, Jackson AA, Elia M, et al. Fetal programming of body composition: relation between birth weight and body composition measured with dual-energy X-ray absorptiometry and anthropometric methods in older Englishmen. Am J Clin Nutr 2005;82:980-7.ArticlePubMed
• 39. Symonds ME, Pope M, Sharkey D, Budge H. Adipose tissue and fetal programming. Diabetologia 2012;55:1597-606.ArticlePubMedPDF
• 40. Alser M, Naja K, Elrayess MA. Mechanisms of body fat distribution and gluteal-femoral fat protection against metabolic disorders. Front Nutr 2024;11:1368966.ArticlePubMedPMC

Figure & Data

References

Citations

Citations to this article as recorded by  

About this article

Ding D, Luo X, Chen S, Liu Z, Kuang X, Zhuang T, She G, Huang H, Yang X, Li J, An R. Birth Weight, Adult Fat Distribution, and Type 2 Diabetes Mellitus Risk: Sex-Specific Study in a Large Prospective Cohort. Diabetes Metab J. 2026 Jan 29. doi: 10.4093/dmj.2025.0569. Epub ahead of print.

Received: Jun 29, 2025; Accepted: Oct 23, 2025

DOI: https://doi.org/10.4093/dmj.2025.0569.

Download citation ↓