Association between total and regional body fat measured by dual-energy X-ray absorptiometry and apolipoprotein B

Abstract

Apolipoprotein B (apoB) can be measured directly and accurately, and better predicts atherogenic risk than conventional lipid profiles. We aimed to investigate whether total and regional (trunk or leg) fat deposits are associated with apoB levels in general US adults. 4585 participants were enrolled from the US National Health and Nutritional Surveys from 2011 to 2016. Overall and regional body fat were measured using dual-energy X-ray absorptiometry. The associations of total and regional fat with apoB levels were evaluated using linear regression models. Following adjustment for demographic, lifestyle, and clinical risk factors, whole-body fat percentage was positively associated with apoB levels. Additionally, percent trunk fat was positively associated (highest vs. lowest tertile beta = 17.73 for men and 14.89 for women, respectively), whereas percent leg fat was inversely associated (highest vs. lowest tertile beta = − 4.84 for men and − 6.55 for women, respectively) with apoB levels in both sexes. The association for trunk fat and leg fat remained significant after further adjustment for body mass index or waist circumference. Higher percent trunk fat combined with lower percent leg fat was associated with particularly higher apoB. In conclusion, among general US adults, both elevated trunk fat and reduced leg fat are associated with higher levels of apoB. Further research is required to elucidate the underlying pathophysiological mechanisms.

Introduction

Obesity is a public health challenge with unabating rates worldwide¹. Traditionally, body mass index (BMI), a simple metric for general adiposity, has been commonly used in population studies and clinical practice. However, it is often criticized for its inability to differentiate between fat and fat-free mass and reflect fat distribution². Individuals with the same BMI could have considerable variations in the amount and distribution of body fat and therefore display different cardiometabolic risks³.

The biological functions of adipose tissue are location-dependent, with upper and lower body fat exerting contrasting impacts (i.e., detrimental vs. beneficial) on various metabolic processes including glucose regulation and lipid metabolism^4–6. There is mounting evidence that trunk fat mass emerged as a strong predictor of metabolic abnormalities, whereas fat accumulation in the legs may possess beneficial effects on metabolic health^7,8. These evidences highlight the potential importance of regional fat distribution in the development of cardiometabolic diseases.

Dyslipidemia is a major determinant for cardiovascular diseases (CVD)⁹. There is a close link between fat mass, its distribution, and the risk of dyslipidemia, including elevated levels of total cholesterol, low-density lipoprotein (LDL-C), and triglycerides; as well as reduced levels of high-density lipoprotein cholesterol (HDL-C)^10–12. It is increasingly acknowledged that compared to conventional lipid indices, including LDL-C and non-HDL-C, apolipoprotein B (apoB) is a more accurate marker of cardiovascular risk¹³. Actually, circulating apoB levels provide a direct measure of the number of more atherogenic particles in the blood because there is a single apoB molecule present on the surface of all atherogenic lipoproteins, such as low-density, intermediate-density, and very low-density lipoprotein¹⁴. In addition, apoB is far less prone to measurement error than LDL-C or non-HDL-C¹⁵. The 2019 European Society of Cardiology/ European Atherosclerosis Society Guidelines have further stated unequivocally that apoB is the best marker of the atherogenic lipoproteins¹⁶. However, studies assessing total and regional fat accumulation (e.g., upper body vs. lower body) and their correlation with apoB are still lacking. In the current study, using body composition data as defined by dual-energy X-ray absorptiometry (DXA) in the US National Health and Nutrition Examination Survey (NHANES), we sought to investigate the associations of whole-body fat, upper-body (trunk) fat, and lower-body (leg) fat with apoB in U.S. adults.

Methods

Study participants

The NHANES, a major program of the National Center for Health Statistics (NCHS), uses a complex, multistage, probability sampling design to evaluate the health and nutritional status of a representative sample of the U.S. civilian non-institutionalized population. The NCHS institutional review board approved the study protocol, and written informed consent was obtained from all participants. The study was carried out following the rules of the Declaration of Helsinki.

We combined three NHANES cycles (2011–2012, 2013–2014, and 2015–2016) for which apoB and DXA data were available. A total of 12,436 adults aged 18–59 years were included. ApoB were detected in 5245 participants. Participants with missing DXA data (n = 660) were excluded. Finally, 4,585 US adults were included in this study.

Measurements

Data were collected via household interviews and physical examinations at a mobile examination center. A standardized questionnaire was used to collect information on age, sex, race, education level, the ratio of family income to poverty (PIR), lifestyle-related risk factors, and medical history. Race/ethnicity was self-reported as Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black, or other. Education was divided into completion of a high school education or lower and education beyond high school. PIR was defined as the ratio of the midpoint of observed family income category to the official poverty threshold and categorized into three groups (≤ 1.30, 1.31–3.49, and ≥ 3.50)¹⁷. Current smoking was defined as having smoked at least 100 cigarettes in their lifetime and currently smoking cigarettes¹⁸. Current drinking was defined as the consumption of at least 12 alcoholic drinks per year¹⁹. The Global Physical Activity Questionnaire was used to reflect physical activity, and metabolic equivalents (METs) were calculated based on the questionnaire to estimate the average weekly energy expenditure²⁰. Standing height, weight, waist circumference (WC), and blood pressure were measured according to standardized protocols^21,22. After participants sat quietly for 5 min, three consecutive blood pressure readings were obtained. If blood pressure measurement was interrupted or incomplete, a fourth attempt was made. All available readings were used to calculate the mean systolic blood pressure and diastolic blood pressure for each participant²³. BMI was calculated as weight in kilograms divided by height in square meters (kg/m²).

Participants aged 8–59 years underwent whole-body DXA scans by professional radiology technicians using Hologic Discovery Model A densitometers (Hologic, Inc., Bedford, MA, USA). The scanned images were analyzed using Hologic APEX software (version 4.0) using the NHANES body composition analysis option. All DXA scans require quality control and analysis, resulting in the measurement of soft tissues and bones in various parts of the body. The trunk and leg regions excluded both the head and arms and were separated by angled lines defining the pelvic triangle. Total fat mass percentage was calculated as the ratio of whole-body fat mass to total mass. Trunk fat percentage was determined by dividing trunk fat mass by trunk mass, while leg fat percentage was derived by dividing leg fat mass by leg mass. The trunk-to-leg fat ratio was defined as the ratio of absolute trunk fat mass to leg fat mass.

Glycated hemoglobin (HbA1c) levels were measured in whole blood samples using high-performance liquid chromatography. ApoB levels were measured directly using immunoassays²⁴.

Statistical analysis

To account for the complex sampling design, the weighted estimates of the population parameters were computed using the NHANES Analytic and Reporting Guidelines²⁵. Data were analyzed using IBM SPSS Statistics, Version 26 (IBM Corporation, Armonk, NY, USA). A two-tailed P lower than 0.05 was taken to indicate significant difference.

Data were presented as mean ± standard deviation (SD) for continuous variables and proportions for categorical variables. Comparisons between groups were performed using the Student’s t-test for continuous variables and the chi-square test for categorical variables. Linear regression analysis was fitted to obtain the beta (β) and 95% confidence intervals (CIs) of apoB according to body fat parameters used either as z-score continuous variables or as tertiles. Four models were constructed to account for the potential confounders. Model 1 included age, race, educational level, PIR, current smoking, current drinking, and where appropriate, regional fat measures (i.e., mutual adjustment for trunk and leg fat). Model 2 further included physical activity, HbA1c levels, systolic blood pressure, and the use of lipid-lowering medications. Model 3 included terms for model 2, and BMI. Model 4 included terms for model 2, and WC. We further analyzed the joint association of trunk and leg fat with apoB by categorizing both body fat measures by tertiles.

Results

The characteristics of the study participants are summarized in Table 1. In total, 2334 men and 2251 women were included in the study. The mean trunk and leg body fat percentages were 27.6 and 27.7 for men, and 36.5 and 43.0 for women, respectively, which differ significantly between genders. Men were slightly younger, had lower educational attainment, and were more likely to be current smokers and drinkers. Moreover, men had significantly greater WC and apoB levels, whereas women had higher BMI.

Table 1.

Characteristics of study participants.

	Men	Women	P
N	2334	2251
Age, years	38.7 ± 0.3	39.7 ± 0.4	0.023
Race, %			0.002
Mexican American	10.6	9.2
Other hispanic	7.4	6.7
Non-hispanic white	64.0	62.8
Non-hispanic black	9.6	12.5
Other race	8.4	8.8
Education level, %			0.025
< High school	17.4	14.5
≥ High school	82.6	85.5
PIR, %			0.072
≤ 1.3	23.7	26.9
> 1.3–3.5	34.9	34.9
≥ 3.5	41.4	38.2
Current smoker, %	24.6	19.5	0.001
Current drinker, %	86.3	72.5	< 0.001
HbA1c, %	5.54 ± 0.02	5.51 ± 0.02	0.263
Systolic BP, mmHg	120.9 ± 0.5	116.1 ± 0.4	< 0.001
Diastolic BP, mmHg	71.9 ± 0.4	69.4 ± 0.3	< 0.001
MET, min per week	6234.5 ± 273.3	3567.8 ± 174.6	< 0.001
BMI, kg/m2	28.7 ± 0.2	29.4 ± 0.2	0.003
WC, cm	100.0 ± 0.5	96.4 ± 0.5	< 0.001
Use of lipid-lowering medication, %	11.4	8.7	0.052
Total fat, kg	24.8 ± 0.3	30.6 ± 0.4	< 0.001
% total fat	27.2 ± 0.2	38.6 ± 0.2	< 0.001
Trunk fat, kg	12.8 ± 0.2	14.5 ± 0.2	< 0.001
% trunk fat	27.6 ± 0.3	36.5 ± 0.3	< 0.001
Leg fat, kg	8.2 ± 0.1	11.6 ± 0.1	< 0.001
% leg fat	27.7 ± 0.2	43.0 ± 0.2	< 0.001
Trunk-to-leg fat rato	1.52 ± 0.01	1.23 ± 0.01	< 0.001
Apo B, mg/dL	93.3 ± 0.8	88.4 ± 0.6	< 0.001

Data were summarized as the mean ± standard error for continuous variables or as a % for categorical variables. PIR the poverty income ratio, HbA1c glycated hemoglobin, BMI body mass index, WC waist circumference, MET metabolic equivalent of task, apoB apolipoprotein B.

The associations of total and regional body fat percent and apoB in the multivariate-adjusted models among men and women are presented in Table 2. The associations were generally significant in all models. After adjusting for age, race, educational level, PIR, current smoking and current drinking (and mutual adjustment for regional fat measures), whole-body fat and trunk fat were positively, whereas leg fat was inversely associated with elevated apoB levels in both sexes, either using adiposity measurement z-scores as continuous variables or categorized in tertiles. Moreover, the increase in apoB per SD increment of trunk-to-leg fat ratio was 6.69 (95% CI 4.96, 8.42) mg/dL in men and 7.46 (95% CI 5.66, 9.27) mg/dL in women. Further adjustments for physical activity, HbA1c levels, systolic blood pressure and the use of lipid-lowering medications did not materially alter the significance. The associations of trunk fat or leg fat percentage with apoB remained significant after additional adjustment for BMI or WC (Fig. 1).

Table 2.

Associations between body fat and apolipoprotein B among men and women.

	Tertile for body fat			P for trend	Each SD increment
	Q1	Q2	Q3
Men
Total fat mass percentage
Range, %	≤ 24.1	24.2–29.2	≥ 29.3
Model 1	Ref	11.25 (7.86, 14.64)	12.60 (9.17, 16.04)	< 0.001	7.37 (5.35, 9.39)
Model 2	Ref	10.22 (6.45, 13.99)	10.73 (7.23, 14.23)	< 0.001	6.40 (4.57, 8.23)
Trunk fat mass percentage
Range, %	≤ 24.0	24.1–30.4	≥ 30.5
Model 1a	Ref	15.06 (10.93, 19.19)	18.44 (14.01, 22.87)	< 0.001	12.99 (9.74, 16.25)
Model 2a	Ref	14.39 (10.21, 18.57)	17.73 (13.73, 21.73)	< 0.001	12.32 (9.27, 15.38)
Leg fat mass percentage
Range, %	≤ 24.4	24.5–29.5	≥ 29.6
Model 1a	Ref	− 2.46 (− 6.79, 1.87)	− 3.77 (− 8.24, 0.70)	0.095	− 7.28 (− 11.00, − 3.56)
Model 2a	Ref	− 3.28 (− 7.12, 0.56)	− 4.84 (− 9.12, − 0.56)	0.028	− 6.99 (− 10.72, − 3.26)
Trunk-to-leg fat ratio
Range	≤ 1.25	1.26–1.64	≥ 1.65
Model 1	Ref	11.60 (7.72, 15.48)	19.78 (15.87, 23.68)	< 0.001	6.69 (4.96, 8.42)
Model 2	Ref	10.05 (6.23, 13.86)	18.05 (14.18, 21.92)	< 0.001	5.88 (4.30, 7.46)
Women
Total fat mass percentage
Range, %	≤ 36.4	36.5–41.8	≥ 41.9
Model 1	Ref	10.44 (6.74, 14.13)	10.09 (5.85, 14.33)	< 0.001	6.11 (3.96, 8.25)
Model 2	Ref	8.58 (4.60, 12.57)	8.19 (3.83, 12.56)	< 0.001	5.02 (2.81, 7.23)
Trunk fat mass percentage
Range, %	≤ 34.1	34.2–40.8	≥ 40.9
Model 1a	Ref	12.65 (9.77, 15.53)	17.00 (13.01, 20.99)	< 0.001	10.64 (8.73, 12.56)
Model 2a	Ref	12.02 (8.28, 15.75)	14.89 (10.13, 19.66)	< 0.001	9.73 (7.14, 12.33)
Leg fat mass percentage
Range, %	≤ 40.5	40.6–45.9	≥ 46.0
Model 1a	Ref	− 1.44 (− 4.91, 2.02)	− 7.89 (− 12.56, − 3.22)	0.001	− 7.61 (− 11.39, − 3.83)
Model 2a	Ref	− 1.10 (− 4.63, 2.43)	− 6.55 (− 11.68, − 1.42)	0.013	− 6.30 (− 10.47, − 2.13)
Trunk-to-leg fat ratio
Range	≤ 1.03	1.04–1.33	≥ 1.34
Model 1	Ref	8.20 (4.97, 11.43)	16.72 (13.01, 20.43)	< 0.001	7.46 (5.66, 9.27)
Model 2	Ref	6.43 (3.15, 9.71)	13.87 (9.51, 18.24)	< 0.001	6.73 (4.78, 8.68)

Model 1 was adjusted for age, race, educational level, PIR, current smoking and drinking.

Model 2 was adjusted for covariates in Model 1 and was additionally adjusted for physical activity, HbA1c, systolic blood pressure, and use of lipid-lowering medication.

^aTrunk fat mass percentage and leg fat mass percentage were mutually adjusted for each other.

Fig. 1.

Association of body fat parameters with apoB. Results were adjusted for covariates listed for Model 2 in Table 2 and additionally adjusted for BMI (Model 3) or WC (Model 4). BMI, body mass index; WC, waist circumference.

When regional fat measures were jointly evaluated, women who had the highest percent trunk fat and the lowest percent leg fat were found to have particularly higher levels of apoB (β = 17.25, 95% CI 3.99–30.51), when comparing those who were in the opposite extreme tertiles of the two measures. Additionally, men who had the highest percent trunk fat and the lowest percent leg fat also had significantly higher levels of apoB (β = 19.83, 95% CI 12.34–27.326) (Fig. 2).

Fig. 2.

Joint association of trunk and leg fat percentage with apoB. Results were adjusted for covariates listed for Model 2 in Table 2. There was no significant interaction between trunk and leg fat percentages on apoB in both men (P for interaction = 0.53) and women (P for interaction = 0.17). β, beta coefficient; CI, confidence interval.

Discussion

In our analysis of general U.S. men and women, higher total fat mass was associated with higher apoB levels. Furthermore, upper-body and lower-body fat exhibited opposing associations with apoB levels, with higher trunk fat being associated with higher apoB levels and higher leg fat being associated with lower apoB levels, independent of known risk factors. Subjects who had both high trunk fat and low leg fat had particularly higher apoB levels when compared with those in the opposite groups of the two measures, especially in women. To our knowledge, this is the first study on DXA-measured total and regional body fat and apoB levels in a general population.

A few prior population studies have investigated the association between body fat and lipid profiles. A Swedish study of 175 men and 417 women found that abdominal fat mass was positively associated, whereas the ratio of gynoid to total fat mass was negatively associated with hypertriglyceridemia in both genders¹⁰. However, none of the fat depots was significantly related to hypercholesterolemia in this study¹⁰. Another Chinese study including 35,832 adults showed a positive association of visceral fat index with the risk of dyslipidemia and its components, and the risk was more pronounced in men¹². Nevertheless, bioelectrical impedance, which may overestimate visceral fat²⁶, was used in the Chinese study.

However, none of these studies have focused on the association between body fat and apoB. Undoubtedly, trapping of an apoB particle within the arterial wall is the essential initiating event in atherosclerosis²⁷. Hence, the number of apoB particles matters more than the cholesterol they carry because it is the particle itself that enters and is the primary driver of this entrapment within the arterial wall²⁸. A growing body of evidence has demonstrated that apoB is superior to LDL-C and non-HDL-C as a marker of cardiometabolic risk¹³. Additionally, apoB can be measured inexpensively, and more accurately, than LDL-C or non-HDL-C¹⁵. The 2019 European Society of Cardiology/European Atherosclerosis Society Guidelines have already endorsed apoB for routine clinical practice¹⁶. Hence, it is important to expand the existing literature on obesity and cardiometabolic disorders by further linking DXA-assessed body fat to apoB levels in the general population.

A larger WC has been associated with an unfavorable lipid profile²⁹. Notably, trunk fat measures, as compared to WC, could better characterize certain upper-body adipose tissue depositions that are more predictive of adverse cardiometabolic risk profiles, such as visceral fat storage³⁰ and liver fat accumulation³¹. In the present study, we provide the first report that greater DXA-measured trunk fat is associated with higher apoB levels after adjusting for WC. Results from the Women’s Health Initiative Study also demonstrated that trunk fat was associated with an elevated risk of CVD, independent of WC³².

Few studies have explored lower-body fat in relation to lipid profiles. Higher gynoid fat percentages were positively correlated with TC in men, whereas gynoid fat accumulation in women was favorably correlated with TG and HDL-C in the U.S. population¹¹. Many epidemiological studies have indicated an inverse association between hip circumference, a measure of gluteofemoral fat deposition, and the risk of dyslipidemia³³. Nonetheless, because hip and gynoid fat measurements capture only parts of the total leg fat, the association between leg fat deposits and apoB merits further investigation. Interestingly, higher leg fat levels demonstrated opposite beneficial associations with apoB compared to higher trunk fat levels in the present study. It is also worth mentioning that previous studies have shown opposing (i.e. detrimental vs. beneficial) associations of upper- and lower-body fat with other cardiometabolic risk factors such as high blood pressure³⁴ and disturbed glucose metabolism³⁵.

Site-specific associations between body fat depots and atherogenic dyslipoproteinemias are plausible, considering that the upper and lower bodies contain different fat compartments with profoundly divergent biological functions. Accumulated fat in the legs may act as a metabolically ‘inert’ sink, where excess calories are stored to buffer energy surpluses³⁶. For given levels of trunk or whole-body fat, a greater proportion of leg fat may indicate a greater lower body fat reservoir and thus protection from ectopic fat accumulation at undesirable sites³⁰. Additionally, evidence shows that adipocytes in the gluteofemoral region have a lower rate of lipolysis and greater free fatty acid uptake than those in the abdominal region³⁷. If excess energy cannot be sufficiently stored in healthy fat depots such as leg fat, ectopic fat deposition may develop as a compensatory mechanism. Ectopic fat accumulation including excess visceral fat leads to impaired insulin sensitivity and increased release of free fatty acids, with subsequent abnormalities in lipid metabolism³⁸. In addition, adipose tissue dysfunction induces an atherogenic secretion pattern, and visceral adipocytes secrete greater quantities of proinflammatory cytokines including tumor necrosis factor-alpha and interleukin 6, which contribute to endothelial dysfunction⁷.

The strengths of our study include the use of a nationally representative cohort of US adults, multivariate adjustment for confounders, and the use of DXA to measure body fat percentage. However, several limitations should be noted. The study is cross-sectional in design and, therefore, does not allow the inference of causality. Further clinical trials are needed to determine whether improving regional fat distribution can result in apoB reduction. Second, DXA cannot distinguish visceral fat from subcutaneous fat in the trunk region; further studies are warranted to explore their association with apoB individually. Lastly, only adults aged 18–59 years were eligible for DXA scans, and hence the results may not be generalizable to older individuals.

Conclusion

Our study indicated that higher total body fat was associated with higher apoB levels. Moreover, greater trunk fat was associated with higher apoB levels, while greater leg fat was associated with lower apoB levels. These findings underscore the importance of body fat amount and distribution for cardiometabolic health.

Author contributions

H.L., N.W., and Q.C. designed the study, contributed to the discussion, reviewed, and edited the manuscript, and take full responsibility for the entire study. C.C. designed the study, performed the analysis, wrote the manuscript, and contributed to the discussions. C.C. and Y.G. wrote the manuscript, analyzed the data, and contributed to the discussion. J.X., Y.X., C.W., Y.C., H.H., Q.C., and Y.C. analyzed the data and contributed to the discussion.

Funding

This study was supported by the National Key Research and Development Program of the Ministry of Science and Technology of China (2024ZD0532400), the National Natural Science Foundation of China (82100846; 82474398), Shanghai Municipal Health Commission (20234Y0161), Shanghai Administration of Traditional Chinese Medicine (2022QN1001), the Shanghai “Rising Stars of Medical Talent” Youth Development Program (SHWJRS2024-70), and the Shanghai Key Laboratory of Chinese Clinical Medicine (20DZ2272200). The funders played no role in the study design, data collection, management, analysis, interpretation, manuscript preparation, review, or approval.

Data availability

All data analyzed during this study are publicly available on the NHANES website (https://www.cdc.gov/nchs/nhanes/index.htm).

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

This study protocol was reviewed and approved by the National Center for Health Statistics Ethics Review Committee. All participants completed written informed consent forms before participation.