The association between fat distribution and α1-acid glycoprotein levels among adult females in the United States

Abstract

Background

Visceral fat accumulation and obesity-induced chronic inflammation have been proposed as early markers for multiple disease states, especially in women. Nevertheless, the potential impact of fat distribution on α1-acid glycoprotein(AGP), a marker of inflammation, remains unclear. This research was conducted to investigate the relationships among obesity, fat distribution, and AGP levels.

Methods

A cross-sectional observational study was performed using blood samples from adult females recruited through the National Health and Nutrition Examination Survey from 2015 to 2018. Serum levels of AGP were measured using the Tina-quant α-1-Acid Glycoprotein Gen.2 assay. Based on the fat distribution data obtained from dual-energy X-ray absorptiometry assessments, body mass index (BMI), total percent fat (TPF), android percent fat (APF), gynoid percent fat (GPF), android fat/gynoid fat ratio (AGR), visceral percent fat (VPF), subcutaneous percent fat (SPF), visceral fat/subcutaneous fat ratio (VSR) were used as dependent variables. To investigate the link between fat distribution and AGP, multivariate linear regression analysis was utilized. Furthermore, a sensitivity analysis was also performed.

Results

The present study included 2,295 participants. After adjusting for covariates, BMI, TPF, APF, GPF, VPF, and SPF were found to be positively correlated with AGP levels (BMI: β = 23.65 95%CI:20.90–26.40; TPF: β = 25.91 95%CI:23.02–28.80; APF: β = 25.21 95%CI:22.49–27.93; GPF: β = 19.65 95%CI:16.96–22.34; VPF: β = 12.49 95%CI:9.08–15.90; SPF: β = 5.69, 95%CI:2.89–8.49; AGR: β = 21.14 95%CI:18.16–24.12; VSR: β = 9.35 95%CI:6.11–12.59, all P < 0.0001). All the above indicators exhibited a positive dose–response relationship with AGP. In terms of fat distribution, both AGR and VSR showed positive associations with AGP (P for trend < 0.0001). In particular, when compared to individuals in tertile 1 of AGR, participants in tertiles 2 and 3 had 13.42 mg/dL (95% CI 10.66–16.18) and 21.14 mg/dL (95% CI 18.16–24.12) higher AGP levels, respectively. Participants in the highest tertile of VSR were more likely to exhibit a 9.35 mg/dL increase in AGP compared to those in the lowest tertile (95% CI 6.11–12.59).

Conclusions

Overall, this study revealed a positive dose-dependent relationship between fat proportion/distribution and AGP levels in women. These findings suggest that physicians can associate abnormal serum AGP and obesity with allow timely interventions.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12944-024-02223-9.

Introduction

In the United States, 38% of adult women are obese [1]. Obesity, once believed to be just a metabolic abnormality, has been shown to demonstrate mutual causality with non-specific immune responses [2–7]. As a result of obesity, women are more vulnerable to fertility issues in addition to metabolic problems including type 2 diabetes and heart disease [8]. Furthermore, the importance of fat distribution is gaining more attention. Accumulation of fat in the abdominal region is linked to health issues related to obesity and even all-cause mortality [9–11]. Conversely, fat tissue gathering in the lower body (gluteofemoral region) has been linked to protective lipid and glucose profiles, along with a decreased risk of cardiovascular and metabolic diseases in population studies [12, 13]. Understanding the full picture of the correlation between body status and fat distribution is vital for health maintenance. In addition, fat distribution is not uniform between men and women. Men and postmenopausal women often exhibit android obesity [14, 15]. This body habitus is also known as an apple-shaped body, due to increased fat in the trunk while the limbs tend to be thin. Women of childbearing age usually demonstrate a gynoid shape [16]. In other words, their bodies tend to have a pear shape due to enhanced fat deposition in the hip and thighs. Differences between these forms of fat distribution are also related to disease predisposition. Excess fat accumulation in the android region is believed to be linked with a higher likelihood of developing cardiovascular disease, hypertension, hyperlipidemia, insulin resistance, and type 2 diabetes [17], whereas gynoid fat accumulation is linked to a lower likelihood of developing metabolic and cardiovascular conditions [18]. In pre-menopausal women altered fat distribution is crucial since android fat accumulation is correlated with a raised prevalence of female infertility [19].

Systemic and tissue-specific chronic inflammation is a common characteristic of obesity [20]. Many findings have suggested that the chronic inflammation caused by fat accumulation differs according to tissue type and distribution. Lim et al. noted that visceral fat in specific tissues released unique inflammatory mediators [21],while Marial et al. suggested that higher fatty deposits in the trunk and inflammation were positively correlated [22]. In women, the release of IL-6 from gluteal and femoral adipose tissue is significantly lower than that from abdominal subcutaneous fat [23]. Therefore, fat distribution strongly correlates with inflammation, with gynoid fat demonstrating a more beneficial inflammatory profile compared to android fat.

α-1-acid glycoprotein (AGP) is a protein that is produced throughout the body in response to inflammation in the liver and peripheral tissues [24].High levels of AGP are often, frequently indicative of adverse conditions. For example, they can lead to tumor-related immunosuppression [25].However, studies have reported methods to inhibit AGP production. For instance, exercise and some drugs can inhibit AGP production [26–28]. In addition, a good dietary pattern will change the glycosylation of AGP. This change was pointed out to be potentially beneficial [29]. Elevated levels of AGP can be observed in inflammatory patients [30, 31], which have also been reported to be a good indicator of inflammation in patients with polycystic ovary syndrome (PCOS), especially those with infertility [32]. Studies One study, having reported on the association between obesity and AGP, found that the association was stronger in women [33]. Indeed, Prioreschi et al. also demonstrated that fat accumulation was positively associated with AGP levels in South African women [34], further noting [20] that both the trunk/limb ratio and android/ gynoid ratio were positively associated with AGP. What’s more, AGP plays a crucial role in metabolic dysfunction-associated steatotic liver disease (MASLD) which is a well-known inflammation related disease. Studies have shown AGP2 (i.e., ORM2) by activating AMP to effectively hinder adipogenesis which may be a potential target for the treatment of MASLD [35]. Li et al. used pharmacological administration of recombinant AGP2 protein to ameliorate hepatocyte injury and degeneration in mice of MASLD, indicating a complex interaction between AGP and liver health dynamics [36]. More evidence is needed to prove the relationship between AGP and fat distribution, especially in larger populations, to better clarify the association among obesity, fat distribution and inflammatory states. The association between AGP in blood and fat distribution is explained in this study for the first time using data gathered from the National Health and Nutrition Examination Survey (NHANES).

To examine the correlation between obesity and fat distribution in NHANES female participants using dual-energy X-ray absorptiometry (DXA) scans is the objective of this research, which aims to provide worthwhile insights into the health consequences of fat distribution on inflammation and related issues in a wider population.

Methods

Participant selection

Initiated in 1999, the NHANES is an ongoing initiative that evaluates people's nutritional status and general health throughout the United States. Orchestrated by the Centers for Disease Control and Prevention (CDC), this comprehensive survey combines detailed interviews with thorough physical assessments (For more details: http://www.cdc.gov/nchs/nhanes.htm). The interview portion probes into various domains such as demographics, socioeconomic factors, dietary habits, and health-related concerns. Meanwhile, the examination component encompasses a wide range of evaluations including medical and dental check-ups, physiological measurements, and extensive laboratory analyses, all carried out by trained healthcare professionals. The National Center for Health Statistics' Ethics Review Board (Protocol #2011–17 continuation) granted approval for all involved procedures, ensuring that every participant provided written consent prior to participation.

In this study, because AGP data were only available for NHANES survey cycles 2015–2016 and 2017–2018, these cycles were selected. Only women aged 18 to 49 years were included for analysis which encompassed 19,225 participants.

Missing data on AGP; body mass index (BMI); or fat distribution (android percent fat [APF], gynoid percent fat [GPF], visceral adipose tissue mass [VF] and subcutaneous fat [SF]) were excluded. Finally, a total of 2295 participants were included.

Ethical considerations

Each participant gave written consent prior to participation, in accordance with the procedure approved by the Research Ethics Review Board of the National Center for Health Statistics [37]. The NHANES is committed to maintaining strict confidentiality standards and has robust measures in place to safeguard participant anonymity. Given that the current research involved secondary analysis of de-identified data, and that the NHANES dataset is publicly accessible [38], there was no necessity for an institutional review board review for this study [37].

Measurement of AGP

The assessment of AGP was conducted using the Tina-quant α-1-Acid Glycoprotein Gen.2 assay, which operates on the immunological agglutination principle. This process involves the formation of an antigen/antibody complex when anti-AGP antibodies interact with antigens present in the test specimen. This complex leads to agglutination, the intensity of which is quantified turbidimetrically (Refer to: AAGP2 Tina-quant α1-Acid Glycoprotein Gen.2 [package insert]. Indianapolis, IN. Roche Diagnostics. 2014–11, V 9.0.). Each testing sequence included serum quality control (QC) pools from Roche or were generated internally to guarantee accuracy, and was processed in duplicate. These QC samples were then assessed against predefined standards using a robust multi-rule quality control scheme [39]. Data acquisition occurred after the completion of all laboratory analyses. The research team accessed the NHANES database, extracted the pertinent data, and meticulously documented the corresponding measurements.

Measurement of fat distribution

The computation of BMI (kg/m2) involved dividing the individual's weight (in kilograms) by the square of their height (in meters), with the result rounded off to one decimal point. This data collection took place at the Mobile Examination Center (MEC), conducted by skilled health technicians.

As far as body composition analysis goes, DXA is the most widely accepted technique. [40]. Comprehensive DXA scans of the entire body are obtained at the NHANES MEC. When conducting the scanning procedure, the Hologic APEX software is utilized for precise demarcation of the Android and Gynoid (A/G) areas. The bottom trunk section has been specifically labeled as the Android region, with two different lines denoting its borders: a lower line aligning with the pelvic horizontal cut and an upper line automatically positioned above this cut line by the software. The Gynoid region was determined with respect to the Android region's height. The maximum boundary of the Gynoid region is established as 1.5-times the height of the Android region beneath the pelvic line. Conversely, the lower bound of the Gynoid area is set at a distance ensuring the vertical span between the two Gynoid lines is exactly double the height of Android region. The Hologic program carefully placed these demarcating lines to ensure precision and consistency in delineating these essential anatomical regions [41]. The mass of visceral adipose tissue within the abdomen was measured at the approximate level between the L4 and L5 vertebrae. The mass of subcutaneous adipose tissue outside the abdomen was also measured at the approximate level between the L4 and L5 vertebrae. The records of total percent fat (TPF, %), android percent fat (APF, %), and gynoid percent fat (GPF, %) were obtained from NHANES. According to the data for total fat (g), android fat mass (g), subcutaneous fat mass (g), visceral adipose tissue mass (g), and subcutaneous fat mass (g), android fat/gynoid fat ratio (AGR, %), visceral fat/total fat (VPF, %), subcutaneous fat/total fat (SPF, %), visceral fat/subcutaneous fat ratio (VSR, %) were calculated.

Covariates

The study recorded demographic variables such as age, race (Mexican American, Other Hispanic, Non-Hispanic White, Non-Hispanic Black, Other Race), education (Less than High School, High School or GED General Educational Development, Above High School), marital status (Live Alone, Living with a Partner) and the income-to-poverty ratio. Biochemical parameters included serum albumin (g/dL), total cholesterol (mg/dL), triglycerides (mg/dL), and energy intake (kcal). The questionnaire-based variables encompassed disease states such as hypertension or not, high cholesterol level or not, diabetes or not, and lifestyle factors such as physical activity (Vigorous, Moderate, Less Than Moderate) and smoking status.

Data analyses

The analysis followed the recommendation of NHANES on the complex sampling design and weights.

Descriptive statistics were employed for data representation, with continuous variables typically represented by the weighted median and standard deviation (SD), while categorical variables were often depicted using weighted frequency(percentage). The Student 2-tailed t-test or Mann–Whitney U test is utilized to test continuous variables, while the chi-square or Fisher exact test is utilized to test categorical variables.

Applying a multivariate linear regression model to explore the fat distribution's connection with AGP, including an unadjusted model (non-adjusted); a minimally adjusted model (adjust I; adjusted only for age, race, education, marital status, and income-to-poverty ratio); and a fully adjusted model (adjust II; adjusted for age, race, education, marital status, and income: poverty ratio, smoking status, hypertension or not, high cholesterol level or not, diabetes or not and serum albumin, total cholesterol, triglycerides) [42–44]. Calculate β and its 95% confidence intervals (95%CI) to represent the estimated effect value. The tertile of exposure was utilized as an ordinal categorical variable (first to third, with the first tertile set as the reference value) to examine potential trends in this relationship.

To account for missing covariate data, multiple imputation was employed using the R MI procedure, which involved five replications and a chained equation approach. [45, 46], to perform sensitivity analyses (n = 3015).

Data analysis was conducted using R (version 4.3.0; The R Foundation) and Empower (X&Y Solutions Inc) software [47]. The distinction is considered statistically significant at p < 0.05.

Results

Population characteristics

Among those recruited, only 2295 female participants successfully passed all screenings (Fig. 1). The characteristics of participants classified by AGP are reported in Table 1(line 482). These included average age, average AGP level, BMI, total percent fat, android percent fat, gynoid percent fat, AGR, subcutaneous percent fat, visceral percent fat, VSR, total cholesterol, triglycerides, C-reactive protein, and albumin.

Fig. 1.

Flowchart outlining the eligibility and disqualification criteria for female American adults participating in the 2015-2016 and 2017-2018 NHANES of the United States. Abbreviations: AGP: α1 acid glycoprotein; BMI: Body Mass Index; TPF: Total Percent Fat (%); APF: Android percent fat (%); GPF: Gynoid percent fat (%); AGR: Android/Gynoid ratio (%); VPF: Visceral percent fat (%); SPF: Subcutaneous percent fat (%); VSR: Visceral/Subcutaneous ratio (%)

Table 1.

Study population characteristics categorized by α1 acid glycoprotein (AGP)

Variables	Total (n = 2295)	AGP (mg/dL)			P Value
		Low (n = 1004)	Middle (n = 1002)	High (n = 1009)
AGP (mg/dL), Mean ± SD	77.73 ± 24.05	53.13 ± 8.87	75.21 ± 5.73	104.71 ± 16.74	< 0.001
Age (year), Mean ± SD	28.84 ± 11.40	27.39 ± 10.85	29.20 ± 11.49	30.71 ± 11.43	< 0.001
BMI (kg/m2), n (%)					-
< 25	1274 (32.85%)	547 (55.36%)	319 (32.29%)	122 (12.35%)
25–30	1309 (33.75%)	280 (28.06%)	412 (41.28%)	306 (30.66%)
> 30	1295 (33.39%)	128 (12.80%)	298 (29.80%)	574 (57.40%)
TPF (%), Mean ± SD	37.32 ± 6.81	32.87 ± 5.80	37.96 ± 5.74	41.54 ± 5.61	< 0.001
APF (%), Mean ± SD	37.26 ± 8.70	31.78 ± 7.61	37.84 ± 7.54	42.47 ± 6.96	< 0.001
GPF (%), Mean ± SD	41.19 ± 5.52	38.26 ± 5.19	41.68 ± 4.77	43.80 ± 4.83	< 0.001
AGR (%), Mean ± SD	41.11 ± 13.51	34.56 ± 10.80	41.45 ± 12.85	48.31 ± 13.57	< 0.001
SPF (%), Mean ± SD	6.38 ± 0.87	6.22 ± 0.87	6.44 ± 0.89	6.51 ± 0.81	< 0.001
VPF (%), Mean ± SD	1.26 ± 0.53	1.11 ± 0.47	1.27 ± 0.53	1.44 ± 0.55	< 0.001
VSR (%), Mean ± SD	19.62 ± 8.19	17.75 ± 6.98	19.78 ± 8.10	22.15 ± 9.20	< 0.001
Total Cholesterol(mg/dL), Mean ± SD	177.88 ± 37.16	176.45 ± 38.61	176.72 ± 34.75	180.90 ± 38.35	0.012
Triglycerides (mg/dL), Mean ± SD	110.90 ± 75.05	95.63 ± 67.33	112.19 ± 76.93	128.05 ± 76.16	< 0.001
Albumin (g/dL), Mean ± SD	4.15 ± 0.38	4.20 ± 0.41	4.18 ± 0.35	4.07 ± 0.34	< 0.001
Energy intake (kcal), Mean ± SD	1880.80 ± 832.88	1951.37 ± 819.37	1873.27 ± 798.85	1845.20 ± 817.27	0.014
ratio of family income to poverty (%), Mean ± SD	2.28 ± 1.59	2.56 ± 1.65	2.27 ± 1.56	2.08 ± 1.51	< 0.001
Smoke, n (%)					-
Yes	783 (26.06%)	154 (20.18%)	197 (25.99%)	306 (37.55%)
No	2222 (73.94%)	609 (79.82%)	561 (74.01%)	509 (62.45%)
Hypertension, n (%)					-
Yes	456 (13.71%)	67 (7.90%)	115 (13.50%)	168 (18.69%)
No	2869 (86.29%)	781 (92.10%)	737 (86.50%)	731 (81.31%)
High cholesterol level, n (%)					-
Yes	413 (12.42%)	94 (11.08%)	109 (12.79%)	140 (15.59%)
No	2912 (87.58%)	754 (88.92%)	743 (87.21%)	758 (84.41%)
Diabetes, n (%)					-
Yes	151 (3.90%)	18 (1.81%)	44 (4.46%)	61 (6.17%)
No	3718 (96.10%)	974 (98.19%)	943 (95.54%)	927 (93.83%)
Physical Activity, n (%)					-
Vigorous	561 (16.15%)	124 (13.93%)	154 (17.68%)	173 (18.89%)
Moderate	869 (25.02%)	231 (25.96%)	214 (24.57%)	248 (27.07%)
Less than moderate	2043 (58.83%)	535 (60.11%)	503 (57.75%)	495 (54.04%)
Education, n (%)					-
Less than high school	1427 (36.41%)	360 (35.93%)	385 (38.50%)	330 (32.74%)
High school or GED General educational development	667 (17.02%)	144 (14.37%)	166 (16.60%)	202 (20.04%)
Above high school	1825 (46.57%)	498 (49.70%)	449 (44.90%)	476 (47.22%)
Marital status, n (%)					-
Married or living with partner	2116 (63.26%)	560 (66.83%)	514 (60.97%)	548 (61.57%)
Living alone	1229 (36.74%)	278 (33.17%)	329 (39.03%)	342 (38.43%)
Race, n (%)					-
Mexican American	722 (18.39%)	170 (16.93%)	226 (22.55%)	197 (19.52%)
Other Hispanic	427 (10.87%)	94 (9.36%)	136 (13.57%)	95 (9.42%)
Non-Hispanic White	1125 (28.65%)	299 (29.78%)	270 (26.95%)	373 (36.97%)
Non-Hispanic Black	912 (23.22%)	186 (18.53%)	203 (20.26%)	225 (22.30%)
Other Race	741 (18.87%)	255 (25.40%)	167 (16.67%)	119 (11.79%)

The Student 2-tailed t-test or Mann–Whitney U test is utilized to test continuous variables, while chi-square or Fisher exact test is utilized to test categorical variables

The findings demonstrate a statistically significant difference with a P-value of less than 0.05

Abbreviations: AGP: α1 acid glycoprotein; BMI: body mass index; TPF: total percent fat (%); APF: Android percent fat (%); GPF: gynpid percent fat (%); AGR: android/gynoid ratio (%); VPF: visceral percent fat (%); SPF: subcutaneous percent fat (%); VSR: visceral/subcutaneous ratio (%)

The present study confirmed the findings that AGP levels are higher corresponding with higher age and higher BMI. BMI was categorized into three ranges for statistical analysis, and the results showed that BMI > 30 (obese group) had the highest AGP levels (percentage 64.41%). Similarly, fat distribution indicators showed the same trend (all P < 0.05). Experimental results provided clear support that higher BMI and fat mass are associated with higher AGP levels.

Moreover, higher levels of total cholesterol, triglycerides and lower albumin were observed in this population. AGP levels were higher in smokers compared to non-smokers. More detailed data can be found in Table 1.

Multivariate regression analysis

Table 2 demonstrates the correlation between AGP level and fat distribution through the utilization of multivariable linear regression analysis. BMI, TPF, APF, GPF, AGR, VPF, SPF and VSR showed positive correlations with AGP level (all P values < 0.0001) after adjusting for all covariates (BMI: β = 1.31, 95%CI: 1.18–1.45; TPF: β = 1.72, 95%CI: 1.55–1.89; APF: β = 1.30, 95%CI: 1.17–1.43; GPF: β = 1.59, 95%CI: 1.39–1.79; AGR: β = 0.62, 95%CI: 0.53–0.71; VPF: β = 8.58, 95%CI: 5.90–11.25; SPF: β = 2.79, 95%CI: 1.48–4.11; VSR: β = 0.41, 95%CI: 0.26–0.57; all P values < 0.0001). When all the exposures were divided into three quantiles, pronounced dose–response relationships between BMI, TPF, APF, GPF, AGR, VPF, SPF, VSR and AGP levels were observed.

Table 2.

Association between fat distribution and AGP level among American adult female from the National Health and Nutrition Examination Survey 2015–2018

Exposure	Non-adjusted		Adjust Ia		Adjust IIb
	β (95%CI)	P value	β (95%CI)	P value	β (95%CI)	P value
BMI (kg/m**2)
Continuous	1.43 (1.33, 1.52)	< 0.0001	1.37 (1.26, 1.48)	< 0.0001	1.31 (1.18, 1.45)	< 0.0001
Tertile:
13.8–23.6	ref		ref		ref
23.7–30.5	11.97 (10.09, 13.85)	< 0.0001	11.31 (9.09, 13.53)	< 0.0001	9.55 (6.93, 12.17)	< 0.0001
30.6–72.6	26.47 (24.57, 28.36)	< 0.0001	25.75 (23.47, 28.04)	< 0.0001	23.65 (20.90, 26.40)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
TPF (%)
Continuous	1.88 (1.75, 2.00)	< 0.0001	1.93 (1.78, 2.07)	< 0.0001	1.72 (1.55, 1.89)	< 0.0001
Tertile:
15 – 34.4	ref		ref		ref
34.5 – 40.8	14.40 (12.33, 16.47)	< 0.0001	15.41 (12.99, 17.83)	< 0.0001	13.15 (10.38, 15.92)	< 0.0001
40.9—56.1	28.80 (26.73, 30.87)	< 0.0001	29.43 (26.97, 31.89)	< 0.0001	25.91 (23.02, 28.80)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
APF (%)
Continuous	1.43 (1.34, 1.53)	< 0.0001	1.47 (1.37, 1.58)	< 0.0001	1.30 (1.17, 1.43)	< 0.0001
Tertile:
11.9—33.8	ref		ref		ref
33.9 – 42	13.47 (11.46, 15.48)	< 0.0001	14.84 (12.50, 17.18)	< 0.0001	12.70 (10.05, 15.36)	< 0.0001
42.1—58.8	28.39 (26.38, 30.40)	< 0.0001	28.98 (26.65, 31.31)	< 0.0001	25.21 (22.49, 27.93)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
GPF (%)
Continuous	1.82 (1.66, 1.98)	< 0.0001	1.78 (1.60, 1.96)	< 0.0001	1.59 (1.39, 1.79)	< 0.0001
Tertile:
16.6 – 39	ref		ref		ref
39.1 – 43.8	12.22 (10.10, 14.34)	< 0.0001	12.83 (10.42, 15.25)	< 0.0001	10.91 (8.25, 13.58)	< 0.0001
43.9—61.9	23.07 (20.94, 25.19)	< 0.0001	22.50 (20.05, 24.94)	< 0.0001	19.65 (16.96, 22.34)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
AGR (%)
Continuous	0.75 (0.69, 0.81)	< 0.0001	0.75 (0.68, 0.82)	< 0.0001	0.62 (0.53, 0.71)	< 0.0001
Tertile:
15.37 –33.67	ref		ref		ref
33.68 –45.16	13.34 (11.24, 15.43)	< 0.0001	15.25 (12.82, 17.68)	< 0.0001	13.42 (10.66, 16.18)	< 0.0001
45.17—107.47	24.76 (22.67, 26.86)	< 0.0001	26.10 (23.58, 28.62)	< 0.0001	21.14 (18.16, 24.12)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
VPF (%)
Continuous	11.43 (9.65, 13.20)	< 0.0001	12.53 (10.17, 14.89)	< 0.0001	8.58 (5.90, 11.25)	< 0.0001
Tertile:
0.039 – 0.96	ref		ref		ref
0.97—1.39	8.00 (5.66, 10.34)	< 0.0001	9.06 (6.28, 11.84)	< 0.0001	7.80 (4.71, 10.89)	< 0.0001
1.40 – 4.22	14.53 (12.21, 16.85)	< 0.0001	16.72 (13.62, 19.82)	< 0.0001	12.49 (9.08, 15.90)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
SPF (%)
Continuous	3.74 (2.62, 4.86)	< 0.0001	3.75 (2.51, 4.99)	< 0.0001	2.79 (1.48, 4.11)	< 0.0001
Tertile:
3.95 – 6.00	ref		ref		ref
6.01 – 6.75	5.63 (3.25, 8.01)	< 0.0001	6.11 (3.45, 8.77)	< 0.0001	4.89 (2.06, 7.73)	0.0007
6.76—9.36	7.49 (5.11, 9.87)	< 0.0001	7.80 (5.16, 10.44)	< 0.0001	5.69 (2.89, 8.49)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001
VSR (%)
Continuous	0.62 (0.51, 0.73)	< 0.0001	0.65 (0.50, 0.79)	< 0.0001	0.41 (0.26, 0.57)	< 0.0001
Tertile:
0.53 – 15.35	ref		ref		ref
15.36—21.28	3.22 (0.95, 5.49)	0.0055	3.48 (0.79, 6.17)	0.0113	2.69 (-0.30, 5.67)	0.0777
21.29 – 65.83	11.42 (9.16, 13.67)	< 0.0001	12.18 (9.19, 15.18)	< 0.0001	9.35 (6.11, 12.59)	< 0.0001
P for trend		< 0.0001		< 0.0001		< 0.0001

Abbreviations: AGP α1 acid glycoprotein, BMI Body mass index, TPF Total Percent Fat (%), APF Android percent fat (%), GPF Gynoid percent fat (%), AGR Android/Gynoid ratio (%), VPF Visceral percent fat (%), SPF Subcutaneous percent fat (%), VSR Visceral/Subcutaneous ratio (%)

^aModel I: Adjust for: age; race; education; marital status; ratio of family income to poverty

^bModel II: Adjust for: age; race; education; marital status; ratio of family income to poverty; physical activity; energy intake(kcal); smoke status; high blood pressure or not; high cholesterol level or not; diabetes or not; albumin (g/dL); total Cholesterol(mg/dL); triglycerides (mg/dL)

Further analysis revealed a significant correlation: AGP's effect size were observed to be more pronounced in individuals within the second and third quartiles of BMI compared to those in the first quartile (P < 0.0001). This indicates a positive correlation, implying that as AGP level rises, so does the BMI index. Similarly significant positive dose–response relationships were found for both TPF and body fat percent (APF, GPF, VPF, SPF), with quartiles 2 and 3 having meaningfully higher AGP levels than those of quartile 1.

As mentioned above, there was a concentration response relationship between AGR or VSR and AGP. The second and third quantiles of AGR experienced an increase in AGP compared to the first quantile (quantile 2: β = 13.42, 95%CI: 10.66–16.18, P < 0.0001; quantile 3: β = 21.14, 95%CI: 18.16–24.12, P < 0.0001), while the third quantile of VSR also showed an increase in AGP (quantile 3: β = 9.35, 95%CI: 6.11–12.59, P < 0.0001).

Similar to analysis results, the link between AGP and BMI, TPF, APF, GPF, AGR, VPF, SPF, VSR was further confirmed in smooth curve fitting in Fig. 2 which is positive and monotonic. It suggested that the increase in the ratio of android fat to gynoid fat was accompanied by increasing AGP accumulation. In the same vein, the increased ratio of visceral fat to subcutaneous fat was accompanied by an increase AGP accumulation.

Fig. 2.

The red line depicts the fitted smooth curve of the variable, while the space between the two blue lines illustrates the 95% confidence interval (CI)

In sensitivity analysis, the association between AGP and BMI, TPF, APF, GPF, AGR, VPF, SPF, VSR remained robust after the inclusion of participants missing confounders by multiple imputation (Supplemental Table 1).

Discussion

In this study, increased BMI and excess fat accumulation were meaningfully connected with increased α1-acid glycoprotein concentrations in adult females after full adjustment for covariates. Furthermore, in terms of fat distribution, APF, GPF, VPF, and SPF were found to be positively associated with AGP levels. To investigate the influence of different fat distributions on AGP, the android/gynoid ratio and visceral/subcutaneous ratio were taken as research objects and found to show an increasing trend.

There is a multifactorial and singular effect relationship between obesity, inflammation, and chronic disease [2–7]. During chronic inflammation, inflammatory cells such as neutrophils and monocytes infiltrate adipose tissue [48]. Additionally, enlarged adipocytes are more likely to enter a stressed state and release chemokines such as TNF-α and IL-6, which mediate immune cell infiltration [49].Given that gynoid fat distribution is a protective factor in females, it is relatively less likely to cause inflammation. AGP, an abundant human plasma glycoprotein, is an inflammatory marker, whose serum levels can reach up to 5 times during inflammatory events [50]. A different angle on the relationship between obesity and inflammation is provided by the findings, which showed that fat was associated with an increase in AGP.. Consistent with the results, studies have indicated that higher levels of inflammatory markers such as IL-6, TNF-α, and leptin are present in the blood of overweight and obese individuals, including adults and children [51]. Furthermore, Palaniswamy et al. have argued that an increase in BMI and fat accumulation led to increases in inflammatory markers, including AGP [33]. Significantly, these results support their findings. However as much as we know, this is the first anthropometric study in which the risk of obesity and fat distribution on health is evaluated in a population of only women using AGP levels. The relationship between obesity and AGP has been previously studied with data suggesting that obesity is accompanied by an increase in AGP index [33, 52, 53]. Furthermore, Prioreschi et al. emphasized that the accumulation of android fat in South African women is likely the source of an increase in AGP [34];with Black South African women gaining extra fat around their abdomens. The data indicate that the distribution of AGP levels differs according to race. Mexican American and other Hispanic groups had the largest number of people with moderate AGP levels, accounting for 21.18% and 14.74% of the total population, respectively Non-Hispanic Whites accounted for high AGP levels, indicating a greater incidence of obesity among this group. Similarly, the proportion of non-Hispanic Black people with high AGP levels was the largest (22.50%). In contrast, the total number of individuals with low AGP levels was higher in other ethnic groups. One study shown that the prevalence of obesity increased between 2017 and 2018 in both non-Hispanic Whites and non-Hispanic Blacks [54].This data forms the foundation for the viewpoints presented in the research. The variations in obesity levels among ethnic groups may be attributable to differences in sociodemographic status [55].

The present study corroborates findings that the accumulation of visceral and android fat is associated with adverse outcomes and elevated levels of AGP. Consistent with these conclusions, it has been demonstrated that disparities in fat distribution contribute to altered metabolism and an elevated risk of metabolic diseases [56].Android adipocytes tend to increase in size and become more sensitive to lipolytic stimuli. Additionally android fat discharges more lipolysis products into the systemic circulation than gynoid fat. On the contrary, gynoid fat can better retain fatty acids and other lipolysis products, which play a protective role in metabolism [43, 57].Android obesity, also known as abdominal obesity, is harmful to women, since it can lead to abnormal hormone levels [58] and may be related to some forms of female infertility. Abdominal obesity not only causes hypothalamic-pituitary–gonadal axis dysfunction in women with ovulatory dysfunction, but also has toxic effects on reproductive tissue due to excessive fatty acid degradation. This leads to germ cell damage and a chronic low-grade inflammatory state [59–61].Intriguingly, Broughton et al. point out that obese women remain infertile even in the absence of ovulatory dysfunction, and it appears that obesity has an effect on the outcome of assisted reproductive technology [62]. As delineated previously, AGP has been identified as an excellent marker of inflammation in patients with PCOS, particularly in those with concurrent infertility. A promising research direction may be the potential association between female infertility and AGP. In short, the results further clarify the association between adiposity, fat distribution and AGP, consistent with previous findings of others groups. This study provides evidence supporting the association between inflammation-related diseases and variations in fat distribution, suggesting that AGP should be considered as a critical indicator of disease in females.

Compared to subcutaneous fat, visceral fat predominantly comprises larger and dysfunctional adipocytes, and is associated with high levels of fatty acid degradation and adipokine secretion leading to inflammation [63–66]. AGP glycoforms are altered during inflammation, thereby positioning AGP as a potential detection index for certain diseases [67].The findings of these experiments indicate that the accumulation of visceral fat correlates with elevated levels of AGP. Previous studies have demonstrated that fatty acids and cytokines released from visceral fat contribute to insulin resistance [68–70].Due to its unique anatomical location, primary hepatic insulin resistance induced by visceral fat may lead to glucose metabolic dysfunction in patients [71].It was hypothesized that higher levels of AGP were related to changes in the cellular environment of the liver. Additionally, females, as a demographic, exhibit variations in visceral fat production. As previously mentioned, women of childbearing age whose estrogen antagonizes the production of visceral fat [14], therefore, visceral obesity is more usual in men and postmenopausal women. Similarly, visceral fat deposition occurs in women with abnormally high androgen levels [72]. Moreover, estrogen itself reduces AGP synthesis [73]. Thus, it was suggested that the accumulation of visceral fat in women of childbearing age may be due to the effects of unbalanced estrogen activity. In summary, a correlation exists between the accrual of visceral adiposity and elevated levels of AGP. In women the detrimental effects of elevated AGP are linked to the endocrine system which will be the focus of future research.

Study strengths and limitations

Although this study used NHANES data to analyze the correlation between fat distribution and the serum AGP levels for the first time, several limitations should be noted. To begin with, the analysis was conducted only on the adult female population, and the representativeness of the sample needs to be improved. Therefore, whether the findings can be extrapolated to other, larger populations (males, adolescents, and the elderly) needs to be confirmed. Secondly, due to the cross-sectional nature of this study, it is not feasible to definitively determine the causal link between obesity and AGP. There are various reasons for changes in AGP levels, and different conformations of AGP have different physiological functions [74]. It is too simplistic to use AGP as an inflammatory marker and the correlation between obesity and levels of inflammation requires more detailed and diverse indicators. It is worth mentioning that other markers such as C-reactive protein (CRP), adiponectin, inerleukin-6, FGF and TGF [75–79] have been repeatedly reported to have a strong relationship with adiposity and obesity, which also supports this conclusion. Nevertheless, the ability of a single marker to serve as a reliable indicator may be limited. Therefore, developing a combined index of multiple markers as a composite index could represent an innovative approach. Additionally, there are various types of adipose cells, each with its own function and significance [80, 81].To understand the structural and functional aspects of fat cells from different locations, it is essential to conduct animal studies that analyze these cells and their roles at the transcriptome level.

Conclusions

The study established that there is a correlation between fat distribution and AGP levels in adult North American females. Android fat and visceral fat will become our focus in future research. It is essential to prevent the transition to tissue-specific obesity in overweight individuals. It is worth noting that obesity should be considered when abnormal serum AGP levels are detected during physical examinations. Good control of body fat, especially visceral fat, beneficial in improving AGP-related inflammation. Recognizing differences in fat deposition helps to identify obese individuals at risk for inflammation, enabling the implementation of early interventions for those at high risk. The results cast a new light and provide insights into identifying women at risk for poor metabolic health.

Abbreviations

AGP

α1 Acid glycoprotein

AGR

Android fat/gynoid fat ratio

APF

Android percent fat

BMI

Body mass index

CAD

Coronary artery disease

DXA

Dual-energy x-ray absorptiometry

GPF

Gynoid percent fat

MEC

Mobile Examination Center

MASLD

Metabolic dysfunction-associated steatotic liver disease

NHANES

National Health and Nutrition Examination Survey

PCOS

Polycystic ovary syndrome

QC

Quality control

SF

Subcutaneous fat

SPF

Subcutaneous percent fat

TPF

Total percent fat

VPF

Visceral percent fat

VF

Visceral adipose tissue mass

VSR

Visceral fat/subcutaneous fat ratio

Authors’ contributions

Each author contributed to this research work in various ways. The contributions of each author are outlined below: FQ designed the study and performed data screening; TZ conducted bioinformatics analysis of the data; YT and SW were responsible for writing the manuscript and organizing the figures; MW handled manuscript revisions and data verification; DW provided manuscript editing and guidance.

Funding

This work was supported by the National Natural Science Foundation of China (no. 82160379 and no.82072195) and Collaborative Innovation Center of Chinese Ministry of Education (2020–39).

Availability of data and materials

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.