Abstract
Background and objective
A Body Shape Index (ABSI) serves as a potential indicator of fat distribution, offering a more reliable association with all-cause mortality compared to overall adiposity. The present cohort study aims to explore the relationship between ABSI and all-cause mortality in US adults with Type 2 Diabetes (T2D).
Methods
For this cohort study, we extracted information on 5,461 US adults with T2D from the National Health and Nutrition Examination Survey (NHANES) and the NHANES Linked Mortality File. Trends in ABSI from 1999 to 2018 were calculated and analyzed using partial Mann–Kendall tests. To assess the relationship between ABSI and all-cause mortality, as well as the robustness of the association results, we employed weighted restricted cubic splines (RCS), weighted Cox proportional hazards models, sensitivity analyses, and stratified analyses. Additionally, we conducted time-dependent receiver operating characteristic (ROC) curve analysis to evaluate ABSI’s predictive capability for all-cause mortality over 3, 5, and 10 years.
Results
Among US adults with Type 2 Diabetes (T2D), the mean ABSI gradually increased from 0.08333 to 0.08444 between 1999 and 2018. Following a median follow-up period of 90 months, 1,355 deaths (24.8% of the participants) occurred due to all causes. A left J-shaped association was observed between ABSI and all-cause mortality, with a 39% increased risk among US adults with T2D who had an ABSI below 0.08105 after full adjustment.
Conclusion
Our research has demonstrated a significant association between an elevated ABSI and the risk of all-cause mortality among US adults with T2D. These findings support the potential use of ABSI as a noninvasive tool to estimate mortality risk among US adults with T2D.
Supplementary Information
The online version contains supplementary material available at 10.1007/s40200-025-01570-3.
Keywords: Anthropometry, Body composition, Obesity, Diabetes mellitus, Risk factors
Introduction
Over the last century type 2 diabetes (T2D) is the most prevalent form of diabetes, accounting for 90% of all cases of diabetes [1, 2], among whom 65.3% also exhibit obesity/overweight [3]. Research has demonstrated that fat, particularly visceral fat, is closely associated to insulin resistance [4, 5], various macrovascular and microvascular complications of diabetes such as diabetic nephropathy (DKD) [6, 7], diabetic retinopathy (DR) [8], diabetic peripheral neuropathy [9], atherosclerosis, diabetic foot [9, 10], and other conditions, leading to higher mortality in diabetic individuals [11]. Furthermore, a cohort study by Lee et al. [12] investigated the association between the location of adipose tissue deposits, such as visceral fat area and subcutaneous fat area in individuals at a single university-affiliated healthcare center and all-cause mortality, finding it to be more reliably correlated than the total body fat content. In this context, the identification of individuals with higher visceral fat levels is of crucial significance for planning rational healthcare resources in diabetes management.
The body mass index (BMI), which is calculated by dividing a person’s weight in kilograms by their height in meters squared, has been suggested as a measure to assess obesity. However, the primary limitation of BMI is its inability to differentiate between various body components and accurately assess fat distribution, particularly visceral fat [13, 14]. For instance, Khan et al. [15] indicated that the fat distribution and body composition of individuals with the same BMI, varied significantly. Similar limitations are present with waist circumference (WC), as it does not consider a person’s height and weight, which can result in overestimating or underestimating obesity in individuals of varying body sizes [16]. Computed Tomography (CT), widely regarded as the gold standard for measuring visceral fat, is expensive and necessitates specialized expertise and equipment. Dual-energy X-ray absorptiometry, another technique for assessing body composition, poses potential health risks and is advised to be conducted no more than twice annually [17].
Recently, A Body Shape Index (ABSI), which considers BMI, waist circumference, and height, has emerged as a new anthropometric measure that more accurately reflects fat distribution [11, 18–20]. Numerous studies have demonstrated that ABSI has a stronger association with mortality compared to BMI or WC [11, 12, 19, 21]. To our best knowledge, no national study has examined the link between ABSI and all-cause mortality in US adults with diabetes. To fill these research gaps, we gathered data from a nationally representative sample of U.S. adults aged 20 and older diagnosed with T2D. Our aim was to determine whether higher ABSI values elevate the hazard rate of all-cause mortality and to identify the probable ABSI cut point that signals a high risk of mortality among T2D patients.
Methods
This cohort study collected de-identified, publicly available data from the National Health and Nutrition Examination Survey (NHANES), which is a periodic, cross-sectional sampling survey conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention and was approved by the National Center for Health Statistics Institutional Review Board. Hence, it was deemed exempt from ethical review and informed consent by the Capital Institute of Pediatrics Academic Review Board. This study was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cohort studies.
Study population
All participants in this study were drawn from the National Health and Nutrition Examination Survey (NHANES), a nationally representative sample. The information gathered encompassed demographic and socioeconomic characteristics, health-related behaviors, and health conditions. During study recruitment, trained interviewers administered and collected standardized questionnaires from the non-institutionalized US civilian population. Trained medical professionals conducted physical measurements and laboratory tests at mobile examination centers. For details on the sampling methods and analytical guidelines utilized in this study, please refer to the Centers for Disease Control and Prevention’s (CDC) publication on NHANES survey methods and analytical guidelines [22].
In this cohort study, we included eligible participants from 10 cycles of NHANES from 1999 to 2018. We defined T2D based on participants meeting any one of the American Diabetes Association criteria: (1) self-reported physician diagnosis of diabetes; (2) receipt of oral glucose-lowering medicines or insulin; and (3) fasting plasma glucose level of at least 126 mg/dL, 75-g oral glucose tolerance test of at least 200 mg/dL (to convert glucose to millimoles per liter, multiplied by 0.0555), or hemoglobin A1c (HbA1c) level of at least 6.5% (48 mmol/mol) (to convert HbA1c percentage of total hemoglobin to a proportion of total hemoglobin, multiplied by 0.01). Participants were identified as pregnant or having cancer based on their responses to questions from the questionnaire component of NHANSE. The questions were: "Are you pregnant now?" and "Have you ever been told by a doctor or other health professional that you had cancer or a malignancy (ma-lig-nan-see) of any kind?".
In our study, a total of 9,517 individuals were enrolled, all of whom had type 2 diabetes (T2D) and available data concerning ABSI and final mortality status. At baseline, participants were excluded for a variety of reasons, with the specific details pertaining to these exclusions outlined in eFigure 1 of supplement. Subsequently, the final analytical cohort consisted of 5,461 patients with T2D.
Measurement of ABSI
Information on the height (cm), weight (kg), and WC (cm) of the included individuals was directly obtained from the 1999 to 2018 NHANES. Height (cm) and weight (kg) were measured exactly to two decimal places, and WC (cm) was measured midway between the lower rib margin and the iliac crest in a standing condition and measured exactly to one decimal place. Based on the height (cm), weight (kg), and WC (cm), ABSI was calculated as follows: ABSI = WC / (BMI2/3 × height1/2)18.
Ascertainment of mortality
Our study obtained mortality data ascertained through December 31, 2019 from the Centers for Disease Control and Prevention (CDC) website [23]. Follow-up time was calculated from the date of survey participation to the date of death or the end of follow-up, whichever occurred first.
Assessment of covariates
Questionnaires were used to obtain information on sociodemographic characteristics, smoking status, alcohol consumption, physical activity, total energy intake, diabetes duration, diabetes medication use, hypertension, and hypercholesterolemia. Educational attainment was categorized as less than high school, high school, or some college or more [1]. The family income-poverty ratio level was defined as the total family income divided by the poverty threshold and classified as less than 1, 1 to 3, or more than 3 [24]. Smoking status was categorized as never smokers (fewer than 100 cigarettes in their lifetime), former smokers (quit after smoking more than 100 cigarettes), or current smokers [25]. Participants were classified as being a nondrinker, or drinker according to the response to question in questionnaire: “How often drink alcohol over past 12 months?” [26]. Participants were classified as physically active if they had at least 150 min of moderate-to-vigorous physical activity per week; otherwise, they were classified as physically inactive according to the Centers for Disease Control and Prevention Physical Activity Guidelines for Americans [27]. The daily total energy intake was the mean of the participants’ 2 daily total energy intake collected from the dietary recall questionnaire. If participants had only one recall data point, this was used as the daily total energy intake in the study.
In addition, metabolic biomarkers, including HbA1c, total cholesterol, triglycerides, low-density lipoprotein, neutrophils, and lymphocytes, were measured at baseline among participants who provided blood samples. The Neutrophil-to-Lymphocyte Ratio (NLR) was obtained by calculating the ratio of the neutrophil count to the lymphocyte count [22].
Statistical analyses
Given the complex sampling design of the NHANES, we took sample weights into account to represent national non-institutionalized civilian US residents. Sample characteristics are reported as mean ± standard deviation for normally distributed continuous variables, medians (interquartile ranges) for non-normally distributed continuous variables, and numbers with percentages for categorical variables.
Weighted mean ABSI and 95% CIs were calculated and compared across 10 cycles overall, with the trends examined by Mann–Kendall tests. Categorical variables were compared using the survey-weighted chi-square test, while continuous variables were compared using survey-weighted linear regression. Restricted cubic splines (RCS) with five knots were deployed to explore the potential nonlinear relationships between ABSI and all-cause mortality among individuals with T2D. The Wald test was performed to test for nonlinearity. Survival analysis was conducted using the Kaplan–Meier method and the survival probability at different ABSI levels was evaluated, with comparison using the log-rank test. If the Cox model assumption of proportionality was valid, survey-weighted Cox proportional hazards models were employed to assess the independent association between ABSI and all-cause mortality among individuals withT2D. The outcomes are presented as model 1(unadjusted), model 2 (adjusted for age, sex, race), and model 3 (adjusted for age, sex, race, education level, family income-poverty ratio level, smoking status, drinking status, physical activity status, daily total energy intake, diabetes medication use, duration of diabetes, self-reported hypercholesterolemia, self-reported hypertension, and cholesterol). The median of each ABSI quartile was used as a continuous variable in the regression model for linear trend testing. Stratified and interaction analyses were performed considering demographic, socioeconomic, and health-related behavioral characteristics, such as age (< 45, 45–65, 65–80, ≥ 80 years old), race (Mexican American, non-Hispanic white, non-Hispanic Black, or other), family income to poverty ratio (< 1, 1–3, ≥ 3), smoking status (never smokers, former smokers, or current smokers), physical activity (inactive, active), and alcohol consumption (nondrinker or drinker). The ‘timeROC’ package was employed to evaluate the accuracy of ABSI in predicting survival outcomes at various time points. Finally, to assess the robustness of the association results, subgroup analysis was conducted based on age, sex, race, family income to poverty ratio, smoking status and sensitivity analyses were performed by excluding adults with accidental deaths and adults who died within 2 years after participation.
Variables with missing values exceeding 30% were excluded from subsequent analyses, in accordance with established practices to ensure data quality [25]. To mitigate the impact of the remaining missing data, we implemented multiple imputation using the Predictive Mean Matching (PMM) method through the “mi” package in R. Specifically, we generated five imputed datasets, each representing plausible values for the missing data based on the observed data and assuming that the missingness was at random (MAR). The imputation process accounted for the characteristics of the missing data, as detailed in eFigure 2 of the supplementary material. All statistical analyses were performed using R version 3.6.3 (R Foundation for Statistical Computing). Statistical significance was determined by a 2-sided p value less than 0.05.
Results
Baseline characteristics
A total of 5461 participants were eligible for inclusion. Based on the ABSI quartiles, participants were divided into four groups. Compared with the lower ABSI group, individuals in the higher ABSI group demonstrated several notable differences. They tended to be older, with a larger proportion of non-Hispanic white race and lower levels of family income, were more likely to be former smokers, non-drinkers, and physically inactive, have a longer duration of diabetes and history of hypercholesterolemia and hypertension, and exhibit lower low-density lipoprotein, but higher triglycerides, neutrophils, and NLR (Table 1). During a median follow-up period of 90 (interquartile range (IQR), 48.0–135.0) months, 1355 (24.8%) of the participants died.
Table 1.
Baseline characteristics of participants with diabetes by ABSI levels in NHANES 1999 through 2018a
| Characteristics | Total | Lower ABSI (< 0.08596) | Higher ABSI (≥ 0.08596) | P value |
|---|---|---|---|---|
| Participants, No. (%) | 5788 | 3839 (66.3) | 1949 (33.7) | |
| Age, years | 61.00 [51.00, 70.00] | 59.00 [48.00, 66.00] | 67.00 [59.00, 74.00] | < 0.001 |
| Age, years (%) | < 0.001 | |||
| < 65 | 3491 (60.3) | 2683 ( 69.9) | 808 ( 41.5) | |
| ≥ 65 | 2297 (39.7) | 1156 ( 30.1) | 1141 ( 58.5) | |
| Sex | < 0.001 | |||
| Male | 2966 (51.2) | 1854 (48.3) | 1112 (57.1) | |
| Female | 2822 (48.8) | 1985 (51.7) | 837 (42.9) | |
| Race (%) | < 0.001 | |||
| Mexican American | 1238 (21.4) | 857 (22.3) | 381 (19.5) | |
| Non-Hispanic White | 1911 (33.0) | 1046 (27.2) | 865 (44.4) | |
| Non-Hispanic Black | 1505 (26.0) | 1151 (30.0) | 354 (18.2) | |
| Other b | 1134 (19.6) | 785 (20.4) | 349 (17.9) | |
| Educational attainment (%) | < 0.001 | |||
| < High school | 2116 (36.6) | 1324 (34.5) | 792 (40.7) | |
| High school | 1344 (23.3) | 897 (23.4) | 447 (23.0) | |
| Some college or above | 2320 (40.1) | 1615 (42.1) | 705 (36.3) | |
| Family income to poverty ratio (%) | < 0.001 | |||
| < 1 | 1212 (23.1) | 772 (22.1) | 440 (25.1) | |
| 1–3 | 2458 (46.9) | 1607 (46.1) | 851 (48.5) | |
| ≥ 3 | 1572 (30.0) | 1110 (31.8) | 462 (26.4) | |
| Smoking status (%) | < 0.001 | |||
| Never smokers | 2912 (51.8) | 2082 (55.9) | 830 (43.7) | |
| Former smokers | 1887 (33.6) | 1143 (30.7) | 744 (39.2) | |
| Current smokers | 825 (14.7) | 499 (13.4) | 326 (17.2) | |
| Physical activity (%) | < 0.001 | |||
| Inactive | 4548 (78.7) | 2942 (76.8) | 1606 (82.5) | |
| Active | 1232 (21.3) | 891 (23.2) | 341 (17.5) | |
| Alcohol consumption (%) | < 0.001 | |||
| Non-drinker | 1659 (36.0) | 1019 (33.2) | 640 (41.6) | |
| Drinker | 2953 (64.0) | 2054 (66.8) | 899 (58.4) | |
| Total energy intake, kcal (%) | < 0.001 | |||
| < 1454 | 1445 (25.0) | 904 (23.5) | 541 (27.8) | |
| 1454–1940 | 1449 (25.0) | 940 (24.5) | 509 (26.1) | |
| 1940–2561 | 1447 (25.0) | 988 (25.7) | 459 (23.6) | |
| > 2561 | 1447 (25.0) | 1007 (26.2) | 440 (22.6) | |
| Diabetes medication use (%) | 0.007 | |||
| None | 3675 (80.0) | 2345 (78.9) | 1330 (82.3) | |
| Diabetes medication usec | 916 (20.0) | 629 (21.1) | 287 (17.7) | |
| Duration of diabetes, years (%) | < 0.001 | |||
| < 3 | 791 (19.6) | 567 (21.8) | 224 (15.6) | |
| 3–10 | 1326 (32.8) | 905 (34.8) | 421 (29.3) | |
| ≥ 10 | 1920 (47.6) | 1130 (43.4) | 790 (55.1) | |
| Self-reported disease | ||||
| Hypercholesterolemia (%) | 3102 (53.6) | 1974 (51.4) | 1128 (57.9) | < 0.001 |
| Hypertension (%) | 3646 (63.0) | 2350 (61.2) | 1296 (66.5) | < 0.001 |
| HbA1c (%) | 6.80 [6.20, 7.90] | 6.80 [6.10, 8.00] | 6.80 [6.20, 7.90] | 0.615 |
| Cholesterol, mg/dl (%) | 186.00 [158.00, 218.00] | 189.00 [161.00, 219.00] | 180.00 [154.00, 214.00] | < 0.001 |
| Triglycerides | 132.00 [91.75, 193.00] | 130.00 [90.00, 190.00] | 136.00 [97.00, 199.00] | 0.007 |
| LDL | 105.00 [81.00, 131.00] | 109.00 [84.00, 134.00] | 97.50 [76.00, 125.00] | < 0.001 |
| Neutrophils, × 109/L | 59.10 [52.60, 65.20] | 58.40 [52.00, 64.60] | 60.70 [54.20, 66.40] | < 0.001 |
| Lymphocyte | 2.10 [1.70, 2.70] | 2.10 [1.70, 2.70] | 2.00 [1.60, 2.60] | < 0.001 |
| NLR | 27.71 [20.33, 37.59] | 26.86 [20.08, 35.94] | 29.63 [21.35, 40.68] | < 0.001 |
NHANES National Health and Nutrition Examination Survey, HbA1c Glycated hemoglobin A1c, LDL Low-density lipoprotein, NLR Neurophils/Lymphocyte, Continuous variables are described as means (SEs) or medians [interquartile ranges], when appropriate
Categorical variables are presented as numbers (percentages)
a All estimates accounted for noncomplex survey designs, and all percentages were not weighted
b Categorized based on self-report in the NHANES interview. Other race and ethnicity includes other Hispanic, other non-Hispanic, and multi-race individuals
c Use of insulin or drugs to lower blood sugar
Temporal trends of ABSI
The mean ABSI increased from 0.08333 to 0.08444, and overall temporal trends of ABSI among nonpregnant US population aged 20 years or older and T2D US were statistically significant (eTable 1). Using the mean ABSI from the 1999 to 2000 cycle as a reference, the changes ranged from −0.00014 to 0.00064. The temporal trends of the ABSI are presented in Fig. 1 and eTable 1. ABSI trends across NHANES cycles were stratified by sex and age (eTable 2a and eTable 2b in Supplement). By sex, ABSI was higher in men than in women and exhibited an increasing tendency. Generally, the ABSI increases with aging and across cycles (eFigure 3 in the supplement).
Fig. 1.
Trends of Mean ABSI Values in US T2D Adults From 1999 to 2018
Associations between ABSI and all-cause mortality in T2D individuals
RCS analysis revealed a positive association between ABSI and all-cause mortality (p for all = 0.0017, p for nonlinear = 0.24) (Fig. 2). The risk of all-cause mortality was relatively flat until approximately 0.08408 and then started to increase rapidly afterwards; meanwhile, the hazard ratio (HR) was greater than 1. Survival curve analysis showed a significant decrease in the survival rate in the higher ABSI group compared to the lower ABSI group (p < 0.0001) (eFigure 4 in the supplement). The Schoenfeld test was valid (eFigure 5 in the Supplement), and Cox regression analysis demonstrated that the risk for all-cause mortality was statistically significant for Q3 (ABSI, 0.08407–0.08703) and Q4 (ABSI, ≥ 0.08703) compared with Q1(ABSI, < 0.08105) after controlling for confounders of age, sex, and race. Specifically, after full adjustment, adults within Q4 were 39% more likely to die from any cause than those within Q1 (HR, 1.39; 95% CI, 1.04 to 1.80). In the trend test, a significant linear association was observed between ABSI and all-cause mortality (Table 2).
Fig. 2.
The association of ABSI with all-cause mortality among T2D visualized by restricted cubic spline
Table 2.
The relationships between ABSI and all-cause mortality among US adults with T2Da
| Characteristic | Crude model | Model 1 | Model 2 | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| All-cause mortality | ||||||
| Q1 ABSI | Ref | Ref | Ref | |||
| Q2 ABSI | 1.49 (1.17, 1.92) | 0.002 | 1.16 (1.16, 1.50) | 0.239 | 1.07 (0.79, 1.38) | 0.298 |
| Q3 ABSI | 1.97 (1.58, 2.44) | < 0.0001 | 1.37 (1.03, 1.67) | 0.026 | 1.10 (0.77, 1.44) | 0.179 |
| Q4 ABSI | 3.63 (2.94, 4.47) | < 0.0001 | 1.74 (1.39, 2.19) | < 0.0001 | 1.39 (1.04, 1.80) | 0.0002 |
| P for trend | < 0.0001 | < 0.0001 | 0.008 | |||
HR Hazard ratio, 95% CI 95% confidence interval, ABSI A body shape index, Ref reference
a Data on US T2D adults 20 years or older from the National Health and Nutrition Examination Survey Linked Mortality Files, 1999 to 2018
Crude model, unadjusted
Model 1, adjusted for age, sex, race
Model 2, adjusted for age, sex, race, education level, family income-poverty ratio level, smoking status, drinking status, physical activity status, daily total energy intake, diabetes medication use, duration of diabetes, HbA1c, self-reported hypercholesterolemia, self-reported hypertension, cholesterol, NLR
Subgroup analysis revealed a consistent correlation, with no significant interaction (p > 0.05) observed between sex, race, family income to poverty ratio, smoking status, and physical activity (eTable 3 in the supplement). However, the association between ABSI and all-cause mortality showed a significant interaction with alcohol consumption (p < 0.05) (eFigure 6 in the supplement). Therefore, stratified analyses were performed according to alcohol consumption, and the results showed that diabetic people with an ABSI > 0.08703 accompanying alcohol habits had a greater HR of mortality (eTable 4 in the supplement). Sensitivity analyses showed that significant estimations remained consistent after excluding deaths within 2 years after participation (eTable 5a in Supplement) and accidental deaths (eTable 5b in Supplement).
The predictive ability of ABSI for all-cause mortality among T2D
Time-dependent receiver operating characteristic curve (ROC) analysis showed that the area under the curve (AUC) of the ABSI was 0.651, 0.666, and 0.674 for 3-year, 5-year and 10-year all-cause mortality, respectively (eFig. 7). This indicates that ABSI possesses a similarly effective predictive ability for mortality across different periods.
Discussion
This cohort study comprehensively analyzed the association between ABSI and all-cause mortality utilizing multiple methodologies, demonstrating a positive association between ABSI and all-cause mortality among US T2D populations, which remained consistent across subgroup and sensitivity analyses. Moreover, ABSI exhibited effective predictive capability for all-cause mortality over 3, 5, and 10 years.
The obesity paradox represents a counterintuitive phenomenon wherein individuals possessing a higher BMI demonstrate superior health outcomes or reduced mortality rates when compared to those with normal or lower weights [28]. ABSI, which is not influenced by the obesity paradox, represents a novel anthropometric measure derived through specific algorithms based on measurements of height, weight, and waist circumference [29]. The increasing prevalence of overweight and obese individuals with diabetes underscores the significance of predicting the impact of obesity on mortality rates [30]. Recent studies have underscored the critical role of obesity in exacerbating diabetes-related complications, including DKD [6, 7], DR [8], and diabetic foot [31]. Meanwhile, an elevated obesity-related ABSI has been correlated with increased mortality in aortic-related diseases, as well as an increased risk of diabetes and kidney disease [32, 33]. The evidence suggests that ABSI could potentially evolve into a prognostic marker for predicting disease progression and mortality risk in diverse populations. ABSI has shown predictive capability for mortality [34]. Nonetheless, its potential in the context of T2D remains unclear.
Notably, our study demonstrated a consistent upward trend in ABSI over nearly two decades, consistent with Zhang et al.’s findings on the temporal trends of BRI among U.S. adults aged 20 and older from 1999 to 2018 [26] and subsequent gender-stratified analysis showed that men typically exhibited higher ABSI values compared to women, contrary to Zhang et al.’s previous observations [26]. This discrepancy may be attributed to population variability and differing measures of visceral fat. In the future, cross-sectional studies could be designed to investigate the variations in ABSI distribution across different populations.
Our study suggests that individuals with a higher ABSI in T2D tend to be older, male, non-Hispanic white, have had diabetes for a longer duration, and are more likely to have hypercholesterolemia and hypertension. Both our study and Zhang et al. [26] found that the older participants get, the more visceral fat they possess. A cross-sectional study on T2D found a significant association between ABSI and both age and fat mass and participants with high ABSI scores were notably older and had lower free-fat mass values compared to those with low ABSI scores [20]. Aligning with study of Krakauer et al. [18], ABSI, which is associated with the non-Hispanic white race, showed a consistent increase from midlife into older age and remained consistently higher in males compared to females beyond young adulthood. Consistent with the previous research by Misuzu et al. [35], individuals with higher ABSI are more likely to have diabetes, hypertension, and hypercholesterolemia, highlighting the importance of managing these underlying conditions.
Our research results showed significant disparities in all-cause mortality between ABSI groups when ABSI was used as a categorical variable or continuous variables, suggesting that ABSI, as a potential measures of central adiposity, can be independent predictors of mortality in T2D, consistent with prior research by Li et al. [36] and Emanuela et al. [37]. Another study conducted by Tao et al. [38], which utilized the weight-adjusted waist circumference index to investigate diabetics, reached a similar conclusion. However, compared to the weight-adjusted waist circumference index, previous research on ABSI has already demonstrated its correlation with specific fat distribution and while the weight-adjusted waist circumference index considers the influence of body weight, it fails to consider the impact of height.
Meanwhile, RCS analysis revealed a positive association between ABSI and all-cause mortality. Importantly, our study revealed that an ABSI exceeding 0.08408 is linked to a sharp rise in all-cause mortality. Diabetes patients who maintain an ABSI below 0.08408 were recommended according to the research. The study by Zhang et al. [26] found that the relationship between the BRI—a comprehensive measure of obesity, particularly visceral obesity—and all-cause mortality among noninstitutionalized civilian residents in the U.S. exhibited a U-shaped pattern. A comprehensive study conducted in China discovered that being overweight or obese considerably increased the risk of all-cause mortality in diabetic patients, whereas being underweight lowered the risk of death, although this increase was not statistically significant [36].In contrast to their research, our study revealed a left J-shaped relationship between ABSI and all-cause mortality. Firstly, the body shape indicator we employed differed from those used in the studies. Secondly, fat can serve as a metabolic reserve for older patients, helping to prevent frailty, malnutrition, and osteoporosis [39]. Adipose tissue, an essential component of the endocrine system, can secrete numerous potentially bioactive molecules. These adipokines are vital in regulating energy metabolism, fat distribution, adipocyte function, glucose and lipid metabolism, insulin sensitivity, and the chronic inflammatory response [40, 41]. In conclusion, excessive reduction of fat can impact various metabolic systems. However, obesity also poses health risks, serving as a risk factor for conditions such as hypertension, diabetes, stroke, and cancer [42–44]. Our study suggests that maintaining an ABSI below 0.08408 is advisable, and future cohort studies are needed to establish the optimal ABSI range for reducing mortality in older adults with T2D.
Subgroup analyses revealed an interaction between ABSI and alcohol consumption. In contrast to the findings of Xu et al. [45], which indicated that countries with high diabetes mortality rates typically experienced lower levels of alcohol consumption, our study discovered that alcohol consumption elevates the risk of all-cause mortality in individuals with diabetes. Therefore, it is worthwhile to investigate whether alcohol consumption contributes to an increased risk of all-cause mortality in people with T2D by raising ABSI.
Our study had several strengths. Firstly, the trends of ABSI from 1999 to 2018 as well as the association of ABSI and all-cause mortality were evaluated for the first time. Secondly, employing sampling weight methods improves the extent to which our findings can be generalized to the American population. Notably, our research unveils probable cut point of ABSI to avoid higher all-cause mortality rate among T2D. Methodologically, the study’s results remained consistent even after thorough adjustments for various demographic and disease-related variables.
However, this study has certain limitations. Firstly, the range of confounding factors adjusted for in our analysis might not be comprehensive, leaving room for potential confounders that could impact the relationship between ABSI and mortality. Secondly, since the original data of this study were derived from NHANES, further research is necessary to verify whether the conclusions are applicable in other countries. Additionally, the study did not include comparisons with other anthropometric measures, such as the BRI [46], and did not analyze the difference between them. Importantly, the study did not analyze ABSI’s effectiveness in different populations except diabetic person and further validation in clinical settings.
Conclusion
Our research indicates a substantial and independent link between increased ABSI and the risk of all-cause mortality among individuals with T2D in the United States. When ABSI surpassed 0.08408, the hazard rate for all-cause mortality significantly escalated among U.S. adults with T2D. Notably, ABSI demonstrates strong predictive power for both short-term and long-term all-cause mortality. These findings imply that ABSI could be a straightforward and potential practical tool for identifying high-risk patients and directing specific interventions.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We highly appreciate the work by participants in the NHANES project.
Author contributions
FC, WLL, JT: Investigation, extracted the data, formal analysis, writing and editing original draft; FC, XX, SJX: editing and revising; XSJ, JFZ, QX, RL, WJW, TJ, MQT: Conceptualization, writing, review and editing. All authors reviewed and approved the manuscript.
Funding
The authors reported there is no funding associated with the work.
Data availability
The data used for these analyses are all publicly available at online (https:// www.cdc.gov/nchs/nhanes/index.htm).
Declarations
The study was approved by the National Center for Health Statistics NCHS Research Ethics Review Board. All participants signed the informed consent before participating in this survey.
Competing interests
The authors declare that they have no competing interest.
Footnotes
Feng Chen, Xi Xie and Sijia Xia they are co-first authors.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Feng Chen, Xi Xie and Sijia Xia contributed equally to this work.
References
- 1.Federation ID. Type 2 diabetes. 2024. Available at: https://idf.org/about-diabetes/types-of-diabetes/type-2/. Accessed 25 Jul 2024.
- 2.Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature. 2001;414(6865):782–7. [DOI] [PubMed] [Google Scholar]
- 3.Imbeault P, Haman F, Blais JM, et al. Obesity and type 2 diabetes prevalence in adults from two remote first nations communities in northwestern ontario. Canada J Obes. 2011;2011: 267509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Neeland IJ, Ayers CR, Rohatgi AK, et al. Associations of visceral and abdominal subcutaneous adipose tissue with markers of cardiac and metabolic risk in obese adults. Obesity (Silver Spring). 2013;21(9):E439–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Almurdhi MM, Reeves ND, Bowling FL, Boulton AJ, Jeziorska M, Malik RA. Reduced Lower-Limb Muscle Strength and Volume in Patients With Type 2 Diabetes in Relation to Neuropathy, Intramuscular Fat, and Vitamin D Levels. Diabetes Care. 2016;39(3):441–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Neeland IJ, Ross R, Després JP, et al. Visceral and ectopic fat, atherosclerosis, and cardiometabolic disease: a position statement. Lancet Diabetes Endocrinol. 2019;7(9):715–25. [DOI] [PubMed] [Google Scholar]
- 7.Kurozumi A, Okada Y, Arao T, Tanaka Y. Excess Visceral Adipose Tissue Worsens the Vascular Endothelial Function in Patients with Type 2 Diabetes Mellitus. Intern Med. 2016;55(21):3091–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yinan S, Strong R, Xiaojing Y. The role of visceral fat pigment in human retinal pigment epithelial cells in a high-glucose environment. Int J Ophthalmol. 2020;20(12):2028–33. [Google Scholar]
- 9.Anan F, Masaki T, Yonemochi H, et al. Abdominal visceral fat accumulation is associated with the results of (123)I-metaiodobenzylguanidine myocardial scintigraphy in type 2 diabetic patients. Eur J Nucl Med Mol Imaging. 2007;34(8):1189–97. [DOI] [PubMed] [Google Scholar]
- 10.Matsuzawa Y. Pathophysiology and molecular mechanisms of visceral fat syndrome: the Japanese experience. Diabetes Metab Rev. 1997;13(1):3–13. [DOI] [PubMed] [Google Scholar]
- 11.Sluik D, Boeing H, Montonen J, et al. Associations between general and abdominal adiposity and mortality in individuals with diabetes mellitus. Am J Epidemiol. 2011;174(1):22–34. [DOI] [PubMed] [Google Scholar]
- 12.Lee SW, Son JY, Kim JM, Hwang SS, Han JS, Heo NJ. Body fat distribution is more predictive of all-cause mortality than overall adiposity. Diabetes Obes Metab. 2018;20(1):141–7. [DOI] [PubMed] [Google Scholar]
- 13.Pou KM, Massaro JM, Hoffmann U, et al. Visceral and subcutaneous adipose tissue volumes are cross-sectionally related to markers of inflammation and oxidative stress: the Framingham Heart Study. Circulation. 2007;116(11):1234–41. [DOI] [PubMed] [Google Scholar]
- 14.Huang H, Zheng X, Wen X, Zhong J, Zhou Y, Xu L. Visceral fat correlates with insulin secretion and sensitivity independent of BMI and subcutaneous fat in Chinese with type 2 diabetes. Front Endocrinol (Lausanne). 2023;14:1144834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Khan I, Chong M, Le A, et al. Surrogate Adiposity Markers and Mortality. JAMA Netw Open. 2023;6(9): e2334836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lotfi K, Hassanzadeh Keshteli A, Saneei P, Afshar H, Esmaillzadeh A, Adibi P. A Body Shape Index and Body Roundness Index in Relation to Anxiety, Depression, and Psychological Distress in Adults. Front Nutr. 2022;9: 843155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Marra M, Sammarco R, De Lorenzo A, et al. Assessment of Body Composition in Health and Disease Using Bioelectrical Impedance Analysis (BIA) and Dual Energy X-Ray Absorptiometry (DXA): A Critical Overview. Contrast Media Mol Imaging. 2019;2019:3548284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS ONE. 2012;7(7): e39504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Dhana K, Kavousi M, Ikram MA, Tiemeier HW, Hofman A, Franco OH. Body shape index in comparison with other anthropometric measures in prediction of total and cause-specific mortality. J Epidemiol Community Health. 2016;70(1):90–6. [DOI] [PubMed] [Google Scholar]
- 20.Gomez-Peralta F, Abreu C, Cruz-Bravo M, et al. Relationship between “a body shape index (ABSI)” and body composition in obese patients with type 2 diabetes. Diabetol Metab Syndr. 2018;10:21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ji M, Zhang S, An R. Effectiveness of A Body Shape Index (ABSI) in predicting chronic diseases and mortality: a systematic review and meta-analysis. Obes Rev. 2018;19(5):737–59. [DOI] [PubMed] [Google Scholar]
- 22.Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey (NHANES) survey methods and analytical guidelines. https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx1. October 30, 2021.
- 23.Centers for Disease Control and Prevention. NCHS data linked to mortality files. Available at: https://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm. Accessed October 31,2024.
- 24.Dong G, Gan M, Xu S, Xie Y, Zhou M, Wu L. The neutrophil-lymphocyte ratio as a risk factor for all-cause and cardiovascular mortality among individuals with diabetes: evidence from the NHANES 2003–2016. Cardiovasc Diabetol. 2023;22(1):267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Li Y, Xia PF, Geng TT, et al. Trends in Self-Reported Adherence to Healthy Lifestyle Behaviors Among US Adults, 1999 to March 2020. JAMA Netw Open. 2023;6(7): e2323584. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhang X, Ma N, Lin Q, et al. Body Roundness Index and All-Cause Mortality Among US Adults. JAMA Netw Open. 2024;7(6): e2415051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kany S, Al-Alusi MA, Rämö JT, et al. Associations of “Weekend Warrior” Physical Activity With Incident Disease and Cardiometabolic Health. Circulation. 2024;150(16):1236–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Preston SH, Stokes A. Obesity paradox: conditioning on disease enhances biases in estimating the mortality risks of obesity. Epidemiology. 2014;25(3):454–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Tee JYH, Gan WY, Lim PY. Comparisons of body mass index, waist circumference, waist-to-height ratio and a body shape index (ABSI) in predicting high blood pressure among Malaysian adolescents: a cross-sectional study. BMJ Open. 2020;10(1): e032874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Steele RM, Finucane FM, Griffin SJ, Wareham NJ, Ekelund U. Obesity is associated with altered lung function independently of physical activity and fitness. Obesity (Silver Spring). 2009;17(3):578–84. [DOI] [PubMed] [Google Scholar]
- 31.Li J, Lv JL, Cao XY, et al. Gut microbiota dysbiosis as an inflammaging condition that regulates obesity-related retinopathy and nephropathy. Front Microbiol. 2022;13:1040846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Bawadi H, Abouwatfa M, Alsaeed S, Kerkadi A, Shi Z. Body shape index is a stronger predictor of diabetes. Nutrients. 2019;11(5):1018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lu CY, Chen HH, Chi KH, Chen PC. Obesity indices and the risk of total and cardiovascular mortality among people with diabetes: a long-term follow-up study in Taiwan. Cardiovasc Diabetol. 2023;22(1):345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Krakauer NY, Krakauer JC. Dynamic association of mortality hazard with body shape. PLoS ONE. 2014;9(2): e88793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fujita M, Sato Y, Nagashima K, Takahashi S, Hata A. Predictive power of a body shape index for development of diabetes, hypertension, and dyslipidemia in Japanese adults: a retrospective cohort study. PLoS ONE. 2015;10(6): e0128972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Li CI, Liu CS, Lin CH, Yang SY, Li TC, Lin CC. Association of body indices and risk of mortality in patients with type 2 diabetes. BMJ Open Diabetes Res Care. 2023;11(4):e003474. [DOI] [PMC free article] [PubMed]
- 37.Orsi E, Solini A, Penno G, et al. Body mass index versus surrogate measures of central adiposity as independent predictors of mortality in type 2 diabetes. Cardiovasc Diabetol. 2022;21(1):266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Tao Z, Zuo P, Ma G. The association between weight-adjusted waist circumference index and cardiovascular disease and mortality in patients with diabetes. Sci Rep. 2024;14(1):18973. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Jiang X, Ding L, Guo Y, et al. Association of metabolic syndrome and frailty with postoperative complications in older gastric cancer patients: A body composition perspective. Cancer Med. 2024;13(18): e70194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Han SJ, Boyko EJ, Fujimoto WY, Kahn SE, Leonetti DL. Low Plasma Adiponectin Concentrations Predict Increases in Visceral Adiposity and Insulin Resistance. J Clin Endocrinol Metab. 2017;102(12):4626–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Moon HU, Ha KH, Han SJ, Kim HJ, Kim DJ. The Association of Adiponectin and Visceral Fat with Insulin Resistance and β-Cell Dysfunction. J Korean Med Sci. 2019;34(1): e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Ren Z, Sun W, Wang S, et al. Status and transition of normal-weight central obesity and the risk of cardiovascular diseases: A population-based cohort study in China. Nutr Metab Cardiovasc Dis. 2022;32(12):2794–802. [DOI] [PubMed] [Google Scholar]
- 43.Guan T, Ma J, Li M, et al. Rapid transitions in the epidemiology of stroke and its risk factors in China from 2002 to 2013. Neurology. 2017;89(1):53–61. [DOI] [PubMed] [Google Scholar]
- 44.Nagata N, Sakamoto K, Arai T, et al. Effect of body mass index and intra-abdominal fat measured by computed tomography on the risk of bowel symptoms. PLoS ONE. 2015;10(4): e0123993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Xu Z, Feng J, Xing S, et al. Global trends and spatial drivers of diabetes mellitus mortality, 1990–2019: a systematic geographical analysis. Front Endocrinol (Lausanne). 2024;15:1370489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Thomas DM, Bredlau C, Bosy-Westphal A, et al. Relationships between body roundness with body fat and visceral adipose tissue emerging from a new geometrical model. Obesity (Silver Spring). 2013;21(11):2264–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data used for these analyses are all publicly available at online (https:// www.cdc.gov/nchs/nhanes/index.htm).