{
  "abstract": "Introduction Type 2 diabetes (T2D) is closely associated with excess adiposity, particularly visceral fat. The body roundness index (BRI), calculated from height and waist circumference, provides a refined estimate of visceral adiposity. This study aimed to investigate the associations of baseline BRI and longitudinal changes in BRI with the risk of incident T2D.Research design and methods We used UK Biobank data, a cohort involving adults aged 37–73 years. Both baseline and last follow-up BRI values were classified according to the tertiles of baseline BRI, and long-term BRI changes were categorized by baseline and follow-up grades. The annual average rate of change (AARC) of BRI was also calculated. Cox regression models were employed to evaluate HRs and 95% CIs, while restricted cubic splines explored non-linear relationships.Results Among 485 509 participants, 32 956 developed T2D during the follow-up period. Elevated baseline BRI was correlated with a greater risk of T2D, with adjusted HRs of 2.39 (95% CI: 2.28 to 2.51) and 6.23 (95% CI: 5.96 to 6.50) for the second and third tertiles of BRI, respectively. In the longitudinal analysis of 65 684 participants (1153 T2D cases), low baseline BRI with grade increase was linked to higher risk of T2D (HR: 1.49, 95% CI: 1.05 to 2.13). Among participants with middle baseline BRI, grade decrease and increase showed HRs of 0.65 (95% CI: 0.45 to 0.93) and 1.66 (95% CI: 1.32 to 2.10) versus stable middle. Grade decrease with high baseline BRI was associated with lower T2D risk (HR: 0.50, 95% CI: 0.40 to 0.62) compared with stable high. Non-linear associations between AARC of BRI and T2D risk were identified (p <0.05).Conclusions This study shows that both baseline BRI and long-term BRI changes are linked to T2D risk, emphasizing the significance of monitoring BRI trends for T2D prevention and control.",
  "authors": [
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Xuanli Zhao"
    },
    {
      "affiliations": [
        "Department of Public Health, Shulan International Medical College, Zhejiang Shuren University, Hangzhou, Zhejiang, China"
      ],
      "name": "Fangyuan Jing"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China",
        "Department of Public Health, Shulan International Medical College, Zhejiang Shuren University, Hangzhou, Zhejiang, China"
      ],
      "name": "Yanan Ren"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Jing Zhu"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Xinzhe Jing"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Meiqun Lv"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Ke Huang"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Jing Guo"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Jiayu Li"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Xiaohui Sun"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Yingying Mao"
    },
    {
      "affiliations": [
        "Department of Public Health, Zhejiang Chinese Medical University, Hangzhou, China"
      ],
      "name": "Ding Ye"
    }
  ],
  "full_text": "WHAT IS ALREADY KNOWN ON THIS TOPIC Visceral adiposity is a key risk factor for type 2 diabetes (T2D), and body roundness index (BRI) is a useful marker reflecting body fat distribution.WHAT THIS STUDY ADDS This study links both baseline BRI and its longitudinal changes—including change grades and annual average rate of change—to T2D risk in a large cohort.HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Tracking BRI trends over time may help identify high-risk individuals and guide T2D prevention efforts.Introduction Type 2 diabetes (T2D) is a chronic and complex disorder marked by insulin resistance, reduced insulin production, and elevated blood glucose levels, 1 arising from an interplay of genetic and environmental components. A global disease burden analysis in 2021 projected that the age-standardized prevalence of T2D will increase by 61.2% (95% uncertainty interval: 56.2–68.1), affecting more than 1.27 billion individuals globally by 2050.2 Compelling evidence indicates that obesity and excessive weight gain are major, well-established factors implicated in the emergence and progression of T2D.3 Therefore, identifying and addressing obesity is a critical focus for T2D prevention and management.The body roundness index (BRI) is an innovative anthropometric measure combining height and waist circumference (WC) to estimate body fat composition and assess general health,4 with the ability to more comprehensively reflect visceral fat. It offers a more comprehensive reflection of visceral fat compared with traditional indices such as body mass index (BMI) and WC, addressing the limitation of distinguishing between visceral and overall obesity.4 5 As an emerging anthropometric indicator of body geometry, BRI has demonstrated greater accuracy and efficiency in identifying metabolic components associated with overweight and obesity, particularly in reflecting abdominal adiposity.6 This novel approach enhances the assessment of obesity-related health risks beyond the constraints of conventional measures.Previous studies investigating the correlation of BRI and T2D incidence were limited by a relatively small sample size.7–9 A cohort study of 78 456 individuals, drawn from China’s basic public health service database, reported that prolonged exposure to elevated BRI was correlated with elevated risk from pre-diabetes to diabetes.7 Similarly, a retrospective cohort study of 15 464 Japanese patients also indicated that BRI may be useful as a valid indicator of T2D risk.8 Furthermore, research based on a cohort, which included 9900 elderly Chinese participants, demonstrated that higher BRI levels were positively connected to raised risk of incident T2D in older adults.9 Moreover, most existing studies rely on BRI values assessed at a single time measurement to assess their relationship with outcomes occurring across a long-term follow-up. This static approach fails to capture the shifting patterns of adiposity and its cumulative metabolic consequences. In contrast, evaluating longitudinal changes in BRI offers a more comprehensive understanding of how shifts in body composition influence the risk of developing T2D. The temporal perspective, which better mirrors real-life scenarios than static BRI measurements at a single time point,10 aids in identifying potential causal mechanisms and provides deeper insights into how changes in body composition over time contribute to T2D risk. Interestingly, evidence on the role of temporal changes in obesity indices, including BRI, in predicting T2D remains scarce, underscoring the necessity for further investigation in this area.The study focused on the link between baseline BRI, its long-term change, and the risk of developing T2D. Furthermore, we examined the quantitative changes in BRI and their relationship with T2D incidence. By examining grades of BRI change, we aim to deepen the understanding of the obesity–T2D link and offer practical insights for clinical applications.Research design and methods Study design and participants The UK Biobank is a large, population-based cohort study that enrolled more than 500 000 participants from 22 assessment centers across the UK between 2006 and 2010 at baseline. 11 Comprehensive details about the UK Biobank have been provided in earlier reports.12 Briefly, the study includes three follow-up visits: the first conducted between 2012 and 2013, the second beginning in 2014, and the third commencing in 2019.13As shown in figure 1, a sum of 485 509 individuals with available baseline BRI records, who were free of T2D at enrollment and had not withdrawn from the study, were incorporated into the primary analysis. For the second analysis, 407 389 participants with only one recorded BRI measurement were excluded. Subsequently, 12 436 individuals who were diagnosed with T2D before their last recorded BRI measurement were further excluded. The final analysis comprised 65 684 individuals, aiming to explore how changes in BRI over time relate to T2D risk.Figure 1Flowchart of the study population. BRI, body roundness index; T2D, type 2 diabetes.Assessment of exposure Throughout the interview, trained staff measured participants’ height and WC using standardized procedures. 12 Height was measured with varying precision (to the nearest 0.1 cm, 0.5 cm, or 1.0 cm) based on the assessment center.14 WC was recorded at the midpoint between the iliac crest and lower rib, rounded to the nearest 0.1 cm.15 The BRI was then computed using the following formula BRI=364.2−365.5×(1−[WC(cm)÷2π)]2)÷[0.5×Height(cm)]2, which was developed by Thomas et al.4As the lack of standardized BRI classification, a data-driven approach was adopted, using sample tertiles based on participants with complete baseline BRI records. For those with at least two BRI measurements, both baseline and last follow-up BRI values were categorized according to the same tertile thresholds to ensure consistency. Based on the baseline BRI classification and considering participants’ BRI status at the last follow-up, individuals were further stratified into distinct groups to evaluate longitudinal changes over time.Ascertainment of outcomes Diagnosis codes and the corresponding year were sourced from data category 1712 “First Occurrences,” which captures the earliest recorded diagnoses mapped to the International Classification of Diseases, Tenth Edition (ICD-10) diagnosis based on information consolidated from primary care data, hospital inpatient data, self-reports, and death register records. Incident T2D cases were determined based on the ICD-10 code E11. 16 17 The starting point for calculating person-years of follow-up differed between the two analyses. In the primary analysis, person-years of follow-up were determined starting from the recruitment date until the earliest occurrence of T2D diagnosis, death, loss to follow-up, or the end of the study period (31 October 2022). Differently, for the secondary analysis, follow-up began from the date of the last BRI assessment, while the endpoint remained consistent with those in the primary analysis.Assessment of covariates We controlled for a series of possible confounders, including sociodemographic characteristics, lifestyle behaviors, health status, and medication use. 18–20 Sociodemographic factors included ag, sex, ethnicity, educational attainment, and the Townsend Deprivation Index (TDI), which was divided into quintiles, with higher scores reflecting greater socioeconomic deprivation. Lifestyle behaviors comprised smoking status,21 drinking status,22 physical activity,23 and dietary habits .24 Health status was assessed based on cancer status (ICD-10 codes C00–C97, excluding C44)25 and cardiovascular disease (CVD) status (ICD-10 codes I20–I25, I60–I64, I69).26 In addition, medication use was evaluated, specifically focusing on the use of cholesterol-lowering medications, blood pressure medications, and insulin. To address missing covariate data, multiple imputations were applied to enhance the reliability of the analyses.27Statistical analysis For the study population’s characteristics, variables following a normal distribution were summarized by means±SDs, while skewed variables were described using medians with IQRs. Numbers and percentages were used to report categorical variables. HRs and 95% CIs were calculated using Cox proportional hazards regression models. Model 1 was adjusted for age, sex, ethnicity, education, and the TDI. Model 2 further included adjustments for lifestyle factors, including smoking status, drinking status, physical activity, and diet, in addition to the covariates in model 1. Model 3 incorporated additional adjustments for cholesterol-lowering medication use, blood pressure medication use, insulin use, cancer status, and CVD status. Notably, the analysis examining the association between the annual average rate of change (AARC) in BRI and the risk of T2D was additionally adjusted for baseline BRI, modeled as a continuous variable, to mitigate potential confounding.In the first analysis, baseline BRI was categorized into tertiles (T1, T2, T3) to explore its association with T2D. Then, in the subsequent analysis, BRI change grades were defined based on the baseline and the last recorded BRI measurement. Specifically, participants with baseline BRI in the T1 were categorized into two groups: stable low (ie, T1 to T1) and grade increasing (ie, T1 to T2 or T1 to T3). Those with baseline BRI in the T2 were categorized into stable-middle (ie, T2 to T2), grade-decreasing (ie, T2 to T1), and grade-increasing (ie, T2 to T3) groups. Individuals with a baseline BRI in the T3 were categorized into two groups: stable-high group (ie, T3 to T3) and the grade-decreasing group (ie, T3 to T1 or T3 to T2), which includes those who transitioned from high to either T1 or T2. Importantly, we also computed the AARC in BRI to quantify the degree of change. The AARC was computed by subtracting the baseline BRI from the BRI at the last visit, and then dividing the result by the baseline BRI and the time interval (in years) between the two measurements, ie, (BRIatlastvisit−BRIatbaseline)÷(BRIatbaseline×Years).The exposure–response relationship of BRI and its change with T2D onset was assessed using restricted cubic spline (RCS) regression, which was implemented using three knots set at the 10th, 50th, and 90th percentiles of the exposure distribution. Additionally, stratified analyses were conducted based on sociodemographic characteristics, lifestyle behaviors, medication use, and health status to explore potential effect modifications. Stratified analyses of BRI change grades were not conducted, as the sample size within each subgroup was insufficient to ensure reliable statistical inference. To ensure the stability of our results, we accomplished multiple sensitivity analyses. Primarily, we excluded individuals diagnosed with T2D within 1 year after their last BRI measurement to minimize potential reverse causation. Furthermore, incident T2D cases were identified according to the Field ID of 41270, which represents a summary of the primary diagnosis codes recorded across all hospital inpatient records for each participant. Moreover, in the second analysis, we conducted an additional sensitivity analysis using the baseline as the starting point for follow-up, aligning with the approach of a previous study.28 Lastly, to address potential bias arising from temporal fluctuations in BRI, we conducted a sensitivity analysis focusing on intra-individual variability in BRI trajectories. This analysis was restricted to participants with at least three repeated BRI measurements, allowing for a more robust characterization of longitudinal patterns and reducing the risk of misclassification. For each participant, we calculated the SD of BRI values across all available time points to quantify intra-individual variability. These metrics were then categorized into tertiles.All statistical analyses and visualizations were implemented with R (V.4.3.1). All statistical tests were two sided, and statistical significance was defined as a p value less than 0.05.Results Participants’ characteristics Online supplemental table S1 detailed the baseline characteristics of the study subjects. In the first analysis, the median age of the 485 509 individuals was 58 years (IQR: 50–63 years). Based on BRI tertile classification, increasing trends were observed in median age, the number of participants using medications, and the prevalence of CVD and cancer across tertiles. Among the 65 684 participants who underwent repeated assessments in the following analysis, a total of 1153 T2D cases occurred over the follow-up period. The majority of individuals were categorized into the stable-low (29.7%), stable-middle (18.9%), and stable-high (17.2%) BRI groups (online supplemental table S2). Notably, participants in the stable-high group had a longer median follow-up duration compared with those in the low and middle baseline BRI groups.SP110.1136/bmjdrc-2025-005339.supp1Supplementary dataAssociations of BRI at baseline with incident T2D Of the 485 509 participants, 32 956 developed T2D during a median follow-up of 13.58 years. Participants with higher baseline BRI were significantly associated with an increased risk of developing T2D during follow-up. In tertile-based analyses, the fully adjusted HRs for incident T2D were 2.39 (95% CI: 2.28 to 2.51) and 6.23 (95% CI: 5.96 to 6.50) for the T2 and T3, respectively, compared with the T1 ( table 1). Furthermore, each tertile increase in BRI was linked to a 2.54-fold higher risk of T2D (HR: 2.54; 95% CI: 2.49 to 2.58). RCS revealed a significant J-shaped non-linear correlation of BRI and T2D occurrence (p overall<0.001, p for non-linear<0.001; figure 2). Specifically, when baseline BRI was below 3.93, a decrease in BRI was related to a reduced risk of developing T2D. In contrast, when baseline BRI exceeded 3.93, an increase in BRI was positively correlated with a higher risk of T2D.Figure 2Non-linear relationship between body roundness index (BRI) at baseline with incident risk of type 2 diabetes (T2D).Table 1Associations of BRI at baseline with incident T2DBRITotal/person-yearsEventsModel 1*Model 2†Model 3‡HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P valueT1 (BRI≤3.359)165 490/807 259 15824261.001.001.00T2 (3.359<BRI ≤4.599)163 557/781 948 87275112.63 (2.51 to 2.75)<0.0012.59 (2.47 to 2.71)<0.0012.39 (2.28 to 2.51)<0.001T3 (BRI>4.599)156 462/700 207 62523 0198.19 (7.85 to 8.55)<0.0017.83 (7.50 to 8.17)<0.0016.23 (5.96 to 6.50)<0.001Per tertile increase485 509/2 289 415 65532 9562.96 (2.91 to 3.01)<0.0012.88 (2.83 to 2.94)<0.0012.54 (2.49 to 2.58)<0.001*Adjusted for age, sex, ethnicity, education, and TDI.†Adjusted for age, sex, ethnicity, education, TDI, smoking status, drinking status, physical activity, and diet.‡Adjusted for age, sex, ethnicity, education, TDI, smoking status, drinking status, physical activity, diet, cholesterol-lowering medication use, blood pressure medication use, insulin use, cancers status, and CVD status.T, tertile.BRI, body roundness index; CVD, cardiovascular disease; T2D, type 2 diabetes; TDI, Townsend Deprivation Index.Association between longitudinal BRI change and incident T2D In the second analysis, 1153 cases of T2D were recorded over a median follow-up period of 4.04 years. After stratified by baseline BRI, individuals with low baseline BRI and grade increasing demonstrated a significantly greater risk of developing T2D in comparison to participants in the stable group (HR=1.49, 95% CI: 1.05 to 2.13) ( table 2). Additionally, significant associations were observed for the grade-decreasing group and the grade-increasing group, with HRs of 0.65 (95% CI: 0.45 to 0.93) and 1.66 (95% CI: 1.32 to 2.10), respectively, when contrasted with the stable-middle group. Notably, participants with high baseline BRI and grade-decreasing group showed substantially reduced risk of T2D, with an HR of 0.50 (95% CI: 0.40 to 0.62) compared with the stable-high group. The connection between AARC of BRI and the risk of developing T2D was statistically meaningful in the RCS model (p overall<0.001), with evidence supporting a non-linear relationship (p for non-linearity=0.013; figure 3). The results indicated that when the AARC of BRI was below zero, it was negatively correlated with T2D risk, whereas when the AARC of BRI exceeded zero, it was positively linked to an elevated risk of T2D. Notably, when the AARC of BRI was below −11.37%, the HR dropped below 0.3, while exceeding 25.70% corresponded to an HR greater than 3.Figure 3Non-linear relationship between annual average rate of change (AARC) of body roundness index (BRI) with incident risk of type 2 diabetes (T2D).Table 2Associations of BRI change grades with incident T2DBRI change gradesTotal/person-yearsEventsModel 1* Model 2†Model 3‡HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P valueBaseline BRI: low Stable low19 498/32 913 908781.001.001.00 Grade increasing7778/12 086 200571.71 (1.21 to 2.42)0.0021.69 (1.19 to 2.40)0.0031.49 (1.05 to 2.13)0.026Baseline BRI: middle Stable middle12 424/21 242 4441711.001.001.00 Grade decreasing4856/8 251 392360.60 (0.41 to 0.85)0.0050.61 (0.42 to 0.87)0.0070.65 (0.45 to 0.93)0.020 Grade increasing5546/8 628 9631241.83 (1.45 to 2.31)<0.0011.79 (1.42 to 2.26)<0.0011.66 (1.32 to 2.10)<0.001Baseline BRI: high Stable high11 322/19 308 7995861.001.001.00 Grade decreasing4260/7 461 6791010.46 (0.37 to 0.57)<0.0010.47 (0.38 to 0.57)<0.0010.50 (0.40 to 0.62)<0.001*Adjusted for age, sex, ethnicity, education, and TDI.†Adjusted for age, sex, ethnicity, education, TDI, smoking status, drinking status, physical activity, and diet.‡Adjusted for age, sex, ethnicity, education, TDI, smoking status, drinking status, physical activity, diet, cholesterol-lowering medication use, blood pressure medication use, insulin use, cancers status, and CVD status.CVD, cardiovascular disease.BRI, body roundness index; T2D, type 2 diabetes; TDI, Townsend Deprivation Index.Subgroup analyses Stratified analyses were undertaken to assess whether baseline BRI exhibited differential associations with T2D risk among subpopulations, as presented in online supplemental figure S1–S3. Stratified analyses by sociodemographic characteristics (online supplemental figure S1) indicated that the incidence of T2D was elevated in both the T2 and T3 BRI groups, as well as with each tertile increase, particularly among women, individuals younger than 60 years, and those without a college or university degree (all p<0.001). Significant interactions were observed for age, sex, ethnicity, and education (all p for interaction<0.001), with subgroup heterogeneity present for all subgroups (all p for heterogeneity<0.001). In the stratified analysis by lifestyle factors (online supplemental figure S2), never smokers in both the T2 and T3 BRI groups, as well as with each tertile increase, exhibited a higher risk of developing T2D. Statistically significant interaction effects were observed between BRI and drinking status, as well as between BRI and smoking status, on the risk of incident T2D (p for interaction<0.05). No heterogeneity was identified across these subgroups (all p>0.05). Additionally, interactions were found between baseline BRI and the use of cholesterol-lowering medication, blood pressure medication, baseline cancer status, as well as baseline CVD status, in relation to the incidence of incident T2D (all p for interaction<0.05; online supplemental figure S3). Heterogeneity was also observed across the groups, except for the baseline cancer status group.Sensitivity analyses Consistent associations between baseline BRI and the risk of T2D were observed after excluding individuals who developed T2D in the first years of follow-up, and when restricting the analysis to incident T2D cases identified solely from hospital inpatient records ( online supplemental table S3, S4). Additionally, after excluding participants diagnosed with T2D within 1 year following the last BRI measurement, the previously observed association remained evident (online supplemental table S5). In contrast, no significant association was observed between the grade-decreasing group (HR=0.70; 95% CI: 0.47 to 1.04) and T2D risk in the middle baseline BRI population. Similarly, the association between the grade-decreasing group and T2D risk also lost statistical significance (HR=0.82; 95% CI: 0.60 to 1.11) compared with the stable-middle group in the population with middle baseline BRI, when the analysis was limited to incident T2D cases identified solely through hospital inpatient records (online supplemental table S6). Notably, when the follow-up time was computed from the date of the first recorded BRI measurement, the findings were consistent with the main analysis (online supplemental table S7).In the population with at least three BRI measurements, an increased risk of developing T2D was observed in the SD tertiles (online supplemental table S8). Specifically, the HR for T2D risk was 1.71 (95% CI: 1.01 to 2.89) for T2 and 3.42 (95% CI: 2.12 to 5.53) for T3, compared with T1. Additionally, for each tertile increase in SD, the risk of developing T2D increased by 88% (HR=1.88; 95% CI: 1.49 to 2.37).Discussion To the best of our knowledge, this is the initial large-scale cohort analysis in the UK to investigate the link between BRI and the development of T2D, incorporating the impact of longitudinal BRI changes. Our study identified significant associations between BRI and new-onset T2D. A J-shaped non-linear correlation of BRI and T2D risk was detected through RCS modeling, with an inflection point observed at a BRI of 3.93. Specifically, individuals with a grade increase in the low baseline BRI group demonstrated a raised risk of developing T2D in comparison with those in the stable-low BRI groups. Statistically significant associations were also observed between both the grade-decreasing and grade-increasing BRI and the risk of T2D in the middle baseline BRI population, relative to the stable-middle BRI group. Furthermore, individuals with high baseline BRI in the grade-decreasing group corresponded with a reduced risk of T2D relative to the stable-high BRI group, indicating their potential protective role. Using the RCS model, a non-linear connection was discovered between the AARC of BRI and the risk of T2D. When the AARC of BRI was less than zero, it was negatively associated with T2D risk, whereas a positive correlation was found when the AARC of BRI exceeded zero, suggesting an increased risk of T2D. Specifically, an AARC below −11.37% suggested to be strongly negatively correlated with risk of T2D, while an AARC exceeding 25.70% showed a significantly strong positive association with T2D risk. The results of the sensitivity analysis were generally consistent with the main analysis. Notably, in the population with three or more BRI measurements, the sensitivity analysis revealed that each tertile increase in the SD of BRI was associated with an 88% increase in the risk of T2D. This suggests that the main analysis, which showed a significant association between BRI change and T2D, remains valid even when accounting for the variability in BRI over time.Previous studies have provided supporting evidence for the connection between the BRI and the risk of diabetes, particularly T2D. Data from 15 310 subjects of health examination programs in a retrospective study at Murakami Memorial Hospital revealed that elevated baseline BRI was an independent risk marker for T2D. Furthermore, BRI exhibited a sex-specific non-linear association with T2D risk, with a J-shaped relationship noted in males and inflection points at 3.146 for males and 4.137 for females.29 Analogously, a prospective study based on 10 785 subjects from the China Health and Retirement Longitudinal Study (CHARLS) identified a notable positive link between elevated BRI and greater risk of T2D.30 Moreover, the analysis indicated a non-linear association, showing an inflection at a BRI of 3.96, which aligns closely with our findings. Qiu et al reported a significant non-linear positive association was found between BRI and the occurrence of diabetes and pre-diabetes in US adults.31 Based on a cross-sectional analysis of 11 980 adults aged 20 years and older from the National Health and Nutrition Examination Survey, the results indicate that BRI could be an effective predictor for diabetes and pre-diabetes. An additional cross-sectional investigation carried out in rural northeastern China, which included 5253 men and 6092 women, reported that BRI displayed potential as an alternative anthropometric measure for assessing diabetes risk.32 Given the limited sample sizes of previous studies, our findings further support the correlation between BRI and T2D risk in a substantially larger population-based cohort.However, to our knowledge, there was limited research on the long-term changes of BRI changes and their association with T2D, although studies on other common chronic diseases had been conducted. In recent years, a prospective study based on the Tehran Lipid and Glucose Study cohort investigated 1681 initially healthy women, among whom 320 developed diabetes during follow-up.33 The findings revealed that higher trajectories of BRI showed a significant positive correlation with diabetes risk compared with lower BRI trajectories. Furthermore, two studies from the Kailuan Cohort found that higher BRI trajectories (moderate stable, moderate-high stable, and high stable) were associated with increased CVD risk34 and that a high-stable BRI trajectory was significantly linked to cancer incidence.35 These findings were highly consistent with those from a study involving 9935 participants from the CHARLS, which also found that a higher BRI trajectory was related to an increased likelihood of CVD, suggesting that BRI had potential as a predictive marker for CVD incidence.36 These findings highlighted the importance of dynamic changes in BRI when assessing chronic disease risk. Furthermore, our more granular classification captures trajectories, offering improved risk stratification beyond traditional stable-group categorizations.Although the underlying pathophysiological mechanisms linking BRI to T2D development require additional research, the observed association may be partially explained by the following potential biological mechanisms. Middle or high BRI levels typically reflect greater accumulation of visceral fat, which is metabolically active and releases pro-inflammatory cytokines, including tumor necrosis factor-alpha and interleukin-6, promoting chronic low-grade inflammation. This inflammatory condition impairs insulin signaling pathways, leading to increased insulin resistance.37 38 Indeed, visceral adipose tissue is fundamental to the development of insulin resistance.39 Furthermore, excess visceral fat can directly impair insulin sensitivity in the liver, skeletal muscle, and adipose tissue, all of which are critical target organs for glucose metabolism, causing worsened insulin resistance and a higher risk of T2D.40 Nonetheless, additional research is required to clarify these mechanisms and establish causality.Our study not only investigated the association between baseline BRI and the risk of T2D but also extended the analysis to measure the correlation between long-term BRI change and T2D incidence. By incorporating repeated BRI measurements over time, we were able to identify subgroups characterized by distinct longitudinal changes. Importantly, BRI serves as a simple yet robust anthropometric indicator that better reflects visceral fat accumulation and cardiometabolic risk as opposed to conventional metrics such as BMI and WC. Continuous monitoring of BRI may facilitate early identification of individuals at elevated risk, enabling timely and targeted interventions. Additionally, our findings highlight the public health value of sustained visceral fat reduction through long-term lifestyle modifications to mitigate the growing burden of T2D.Nonetheless, several limitations should be noted. First, the study population is composed exclusively of European individuals, which may restrict the applicability of our results to other ethnic populations. Second, the relatively short follow-up duration in the BRI change analysis may have led to an underestimation of incident cases and attenuated the observed associations. Third, due to data limitations, changes in BRI were assessed using only two time points, and the timing of the last BRI measurement varied considerably across individuals. However, the timing of the last BRI measurement has already been accounted for in the quantitative assessment of change using the AARC metric. This approach helps to mitigate potential bias arising from variations in follow-up duration across individuals and enhances the accuracy of longitudinal evaluations. Furthermore, covariates were adjusted only at baseline due to incomplete follow-up data, possibly overlooking time-varying confounders. Therefore, further studies with extended follow-up periods, more repeated assessments, and more diverse ethnic groups are necessary to validate and extend these findings.In conclusion, this research offers a novel understanding of the link between BRI and T2D. We found that a higher BRI at baseline was strongly associated with an increased risk of developing T2D, with a significant non-linear relationship observed. Furthermore, long-term increases in BRI change were associated with a higher risk of T2D, whereas decreases were linked to a lower risk. These findings point to the value of both the level and longitudinal pattern of BRI as predictive indicators of T2D risk. Continuous monitoring of BRI changes over time could aid in the early detection of individuals at elevated risk and inform targeted interventions aimed at reducing BRI to help prevent the onset of T2D.",
  "title": "Baseline and longitudinal changes of body roundness index and incident type 2 diabetes: evidence from the UK Biobank cohort",
  "uid": "f43126aa-877f-5f5d-bb1a-e611099fdbb4"
}