Funding agencies: This study was supported by a grant from the Ministry of Health, Labour and Welfare of Japan.
Adiponectin and visceral fat associate with cardiovascular risk factors
Article first published online: 11 JUN 2013
Copyright © 2013 The Obesity Society
Volume 22, Issue 1, pages 287–291, January 2014
How to Cite
Matsushita, Y., Nakagawa, T., Yamamoto, S., Kato, T., Ouchi, T., Kikuchi, N., Takahashi, Y., Yokoyama, T., Mizoue, T. and Noda, M. (2014), Adiponectin and visceral fat associate with cardiovascular risk factors. Obesity, 22: 287–291. doi: 10.1002/oby.20425
Disclosure: The authors declared no conflict of interest.
- Issue published online: 11 JAN 2014
- Article first published online: 11 JUN 2013
- Accepted manuscript online: 21 MAR 2013 05:12AM EST
- Manuscript Accepted: 5 FEB 2013
- Manuscript Received: 19 SEP 2012
Objective: To examine the combined effect of CT-measured visceral fat area (VFA) and adiponectin level against the clustering of metabolic risk factors.
Design and Methods: The subjects were 6,996 Japanese. The subjects were divided according to the combinations of VFA and adiponectin level quartiles and the odds ratio for multiple risk factors of metabolic syndrome were calculated (adjusted for age and lifestyle factors using logistic regression analyses). Group with the lowest VFA and the highest adiponectin level was used as a reference. The correlation between adiponectin level and each metabolic risk factor was evaluated.
Results: The strongest correlation was observed between adiponectin level and high-density lipoprotein cholesterol levels (r = 0.369 and 0.439 for men and women). Both VFA and adiponectin level were independently associated with the clustering of metabolic risk factors (interaction P = 0.58 and 0.11 for men and women). The odds ratio for the clustering of metabolic risk factors in the group with the highest VFA and the lowest adiponectin level, compared with the group with the lowest VFA and the highest adiponectin level, was 12.7 (9.7-16.6) for men and 13.5 (6.0-30.2) for women.
Conclusion: The ability to detect metabolic syndrome could be improved by examining adiponectin level in conjunction with VFA.
The prevalence of metabolic syndrome has been growing globally with clusters of obesity, high blood pressure, impaired lipid metabolism, and hyperglycemia. Individuals with metabolic syndrome have a higher risk of cardiovascular disease and a subsequent increase in disease mortality or morbidity . Several criteria for the diagnosis of metabolic syndrome are used worldwide. The visceral adipose tissue is regarded as an endocrine organ, partly because it secretes adipocytokines and other vasoactive substances that can influence the risk of developing traits of metabolic syndrome . We recently demonstrated that measuring the visceral fat area (VFA) is superior in predicting the accumulation of multiple risk factors, compared with the subcutaneous fat area (SFA), BMI, and waist circumference (WC) measurements . Regarding the multiple risk factors of metabolic syndrome, the odds ratios for the VFA quintiles were 1.0 (ref.), 2.4, 3.4, 5.0, and 9.7 for men and 1.0 (ref.), 1.5, 2.6, 4.6, and 10.0 for women (P < 0.001 for trends in both sexes) .
Adiponectin is predominantly secreted by adipocytes, and the adiponectin level is reduced in individuals with obesity, insulin resistance, and type 2 diabetes . Low plasma adiponectin levels have recently been shown to predict the risk of developing type 2 diabetes in humans [9, 11]. The adiponectin level is also inversely associated with other traditional cardiovascular risk factors, such as blood pressure, total and low-density lipoprotein (LDL) cholesterol, and triglyceride (TG) levels [12, 13], and is positively related to high-density lipoprotein (HDL) cholesterol levels [12, 14].
Some previous studies reported the impact of adiponectin levels on metabolic syndrome and its components ; however, the sample sizes were insufficient. In addition, the combined effect of the VFA and adiponectin level has not been examined in an epidemiological study. Thus, we have examined the combined effect of the VFA and adiponectin level on the clustering of metabolic risk factors.
Among the 17,606 employees of the same company and their spouses who underwent a health examination in Japan between 2008 and 2009, we analyzed 6,996 subjects ranging in age from 25 to 75 years (6,221 men and 775 women) who had undergone a computed tomography (CT) examination and answered a questionnaire on lifestyle factors and current treatments for metabolic conditions (hyperlipidemia, hypertension, or diabetes). The VFA was measured using a CT scanner and was calculated using a software application (fatPointer; Hitachi Medico, Tokyo, Japan) according to a protocol described elsewhere . Briefly, single slice imaging at the umbilical level was performed using a CT machine (Redix turbo; Hitachi Medico) while the subject was in a supine position. The imaging conditions were 120 kV, 50 mA, using a 5 mm thick slice. Height, weight, and body fat were measured using an automated scale (BF-220; Tanita, Tokyo, Japan) with the patient wearing a light gown. The BMI was defined as the weight (kg) divided by the square of the height (m2). A blood sample was collected from each subject after more than 12 hours of fasting. The glucose level was measured using the glucose oxidase enzyme-electrode method (A&T, Tokyo, Japan). TG and HDL cholesterol levels were measured using an enzymatic colorimetric method (Cholestest TG; Sekisui Medical, Tokyo, Japan) and a nonsettling enzymatic method (Cholestest NHDL; Sekisui Medical), respectively. Adiponectin levels were measured using an immunoturbidimetric method (Adiponectin Latex Kit for humans; Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan). Blood pressure was measured using an automated sphygmomanometer (Kentaro ADVANCE BP-203RV III A/B; Colin, Tokyo, Japan). This study was approved by the ethics review committee of the National Center for Global Health and Medicine. Written informed consent was obtained from all the subjects.
Definition of the state of risk factor clustering
In this study, subjects were defined using the criteria of the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) guidelines  published in 2005:  high TG (TG ≥ 150 mg/dL),  low HDL cholesterol (HDL cholesterol <40 mg/dL in men and <50 mg/dL in women),  high blood pressure (systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg),  hyperglycemia (fasting glucose ≥100 mg/dL), and  multiple risk factors (having two or more of components [1-4] listed above). Subjects currently receiving treatment for hyperlipidemia, hypertension, or diabetes were deemed as having the respective risk factor, regardless of the biochemical value.
We calculated the Pearson's correlation coefficients between the adiponectin level and each metabolic risk factor. We divided the subjects according to quartiles of the adiponectin level and calculated the odds ratio for multiple risk factors of metabolic syndrome. We adjusted for age, smoking habits (never, past, current), alcohol consumption (nondrinker, drinker consuming less than 2 go per day [the go is a conventional unit of alcohol intake in Japan and contains approximately 23 g of ethanol], or drinker consuming more than 2 go per day), and regular fitness habit (based on a single yes/no question in the questionnaire) using a logistic regression analysis, with the highest adiponectin level group used as a reference. We calculated the odds ratio for the multiple risk factors of metabolic syndrome for a +1 SD increment in the quintile categories of VFA and a +1 SD increment in the quintile categories of adiponectin levels. Furthermore, we divided the subjects according to combinations of VFA and adiponectin level quartiles and calculated the odds ratio for multiple risk factors of metabolic syndrome adjusted for the above-mentioned variables, using the category with the lowest VFA and the highest adiponectin level as the reference. VFA, adiponectin levels, and their interaction term (VFA × adiponectin levels) were included as independent variables in the logistic regression model to examine the interaction effect between VFA and adiponectin levels on the risk of clustering of metabolic risk factors. The stepwise procedure was used to select variables in the multiple logistic regression model with P < 0.1 for entry and P < 0.05 for removal. All the analyses were performed using SPSS for Windows, Version 15.0 (SPSS Inc., Chicago. IL, USA).
The characteristics of the subjects are shown in Table 1. The mean (SD) age of the subjects was 52.8 ± 10.2 years for men and 57.4 ± 9.7 years for women. The mean VFA was 121.7 ± 53.4 cm2 in men and 81.6 ± 46.3 cm2 in women. The mean BMI was 24.1 ± 3.0 kg/m2 in men and 23.0 ± 3.4 kg/m2 in women. The mean log adiponectin level was 0.83 ± 0.20 μg/mL in men and 1.05 ± 0.21 μg/mL in women. The prevalence of multiple risk factors of metabolic syndrome was 45.5% in men and 34.2% in women.
|Body mass index, kg/m2||24.1||(3.0)||23.0||(3.4)|
|Visceral fat area, cm2||121.7||(53.4)||81.6||(46.3)|
|Subcutaneous fat area, cm2||133.6||(56.4)||182.8||(76.9)|
|Adiponectin, log μg/mL||0.83||(0.20)||1.05||(0.21)|
|High blood pressure, %||38.9||34.1|
|High triglyceride, %||35.4||23.7|
|Low HDL cholesterol, %||11.0||17.8|
|Multiple risk factors of metabolic syndrome, %||45.5||34.2|
Table 2 shows the partial correlations between adiponectin level and each metabolic risk factor. The HDL cholesterol level positively correlated with adiponectin level (P < 0.001). Other metabolic risk factors negatively correlated with the adiponectin level (P < 0.001).
The odds ratios for each component of metabolic syndrome according to the adiponectin level are shown in Figure 1. The odds ratios for a high TG level, a low HDL cholesterol level, high blood pressure, and hyperglycemia decreased with increasing quartile categories of adiponectin levels. For the multiple risk factors of metabolic syndrome, the odds ratios (95% confidence intervals [CI]) of the Q1, Q2, Q3, and Q4 adiponectin level categories were 3.4 (3.0-4.0), 2.1 (1.8-2.5), 1.5 (1.3-1.7), and 1.0 (ref.) for men and 4.3 (2.7-6.9), 2.5 (1.6-4.1), 1.5 (0.9-2.4), and 1.0 (ref.) for women. The odds ratio (95% CI) of the lowest (Q1) adiponectin level category for a high TG level was 3.9 (3.4-4.6) in men and that for a low HDL cholesterol level was 5.8 (4.4-7.8) in men and 9.9 (4.8-20.1) in women.
In Table 3, the odds ratios of adiponectin and VFA levels for the clustering of multiple risk factors of metabolic syndrome are shown according to the VFA and adiponectin quartiles, respectively. For men, the increased adiponectin levels were significantly related to reduced clustering of metabolic risk factors, regardless of the VFA category (P = 0.58 for interaction VFA × adiponectin). For women, however, the odds ratio for a +1 SD in the VFA slightly weakened according to the increment of adiponectin level category (P = 0.11 for interaction of VFA × adiponectin). The odds ratio for multiple risk factors of metabolic syndrome according to combined groups of VFA and adiponectin are depicted in Figure 2. The odds ratio for the multiple risk factors of metabolic syndrome in the category with the highest VFA and the lowest adiponectin levels were 12.7 (9.7-16.6) for men and 13.5 (6.0-30.2) for women. We conducted stepwise logistic regression analyses among the largest quartile group of VFA, with hypoadiponectinemia (lowest vs. highest quartile group) as the independent variable. The result revealed that smoking, high TG, low HDL cholesterol, and older age were associated with hypoadiponectinemia in men. Low HDL cholesterol and age were associated with hypoadiponectinemia in women (data not shown).
|n||Odds ratios of +1 SD increment of adiponectin|
|n||Odds ratios of +1 SD increment of VFA|
|Adiponectin (log μg/mL)||≤0.70||1,620||Q1||2.1|
|Adiponectin (log μg/mL)||≤0.91||198||Q1||2.6||(1.7-3.9)|
We observed a strong association between the combined effect of an increasing VFA and a decreasing adiponectin level and the prevalence of multiple risk factors of metabolic syndrome. Both the VFA and the adiponectin level were independently associated with the clustering of metabolic risk factors. Among the components of metabolic syndrome, the adiponectin level had a particularly strong impact on a high TG level in men and a low HDL cholesterol level in both men and women.
Only one previous report, studying 68 obese Korean subjects, demonstrated an association between the adiponectin level and VFA. The adiponectin level was inversely correlated with the VFA (r = −0.691, P = 0.009 in men, r = −0.319, P = 0.002 in women). Levels of a high molecular weight adiponectin also negatively correlated with the VFA (r = −0.650, P = 0.016 in men, r = −0.370, P = 0.005 in women) but not with the BMI or SFA, suggesting hypoadiponectinemia may represent a dysfunction of adipose tissue during obesity .
It was unknown whether adiponectin levels were correlated with disease, even when the VFAs were the same. Therefore, we compared the prevalence of multiple risk factors of metabolic syndrome with combinations of the VFA and adiponectin level. We found a markedly increased risk of clustering of metabolic syndrome among individuals who had a low adiponectin level and a high VFA. It was revealed that even when VFAs were the same, hypoadiponectinemia was associated with older age, smoking, and lipid metabolism (high TG and low HDL cholesterol) in men. In women, hypoadiponectinemia was associated with older age and low HDL cholesterol. Thus, it was confirmed that adiponectin correlated with lipid metabolism independent of VFA; from this, we concluded that adiponectin correlated with the clustering of metabolic risk factors. However, even if the adiponectin levels were the same, weight gain led to a worsening of metabolic risk factors, including lipid. In the largest VFA group, adiponectin itself was related to lipid metabolism and smoking status. Therefore, it was shown that the ability to detect metabolic syndrome would be improved by examining the adiponectin level in conjunction with the VFA, as adiponectin correlated with metabolic risk factors independent of VFA and the correlation was most pronounced between lipid metabolism.
In a case–control study, high plasma adiponectin levels were associated with a lower risk of myocardial infarction (MI) over a follow-up period of 6 years among men without previous cardiovascular disease. After adjustment for matched variables, participants in the highest quintile, compared with the lowest quintile, of adiponectin levels had a significantly decreased risk of MI (relative risk [RR], 0.39; 95% confidence interval [CI], 0.23-0.64; P for trend <0.001). Further adjustment for the hemoglobin A1c or C-reactive protein levels had little impact, but additional adjustment for LDL and HDL cholesterol levels modestly attenuated this association (RR, 0.56; 95% CI, 0.32-0.99; P for trend =0.02) . A multiple logistic regression analysis revealed that hypoadiponectinemia was significantly and independently correlated with coronary artery disease (CAD) (P < 0.0088) among 450 Japanese men. The multivariate-adjusted odds ratios for CAD in the first, second, third, and fourth quartiles (95% confidence) were 2.051 (1.288-4.951), 1.221 (0.684-2.186), 0.749 (0.392-1.418), and 1.000, respectively . Furthermore, another study showed that BMI, serum TG concentration, and the presence of diabetes or CAD remained significantly related to the plasma adiponectin concentration. Weight reduction significantly elevated the plasma adiponectin levels in diabetic obese Japanese subjects (six men and seven women) and nondiabetic obese Japanese subjects (six men and three women). However, the sample size was very small in this study .
This study has several strengths and limitations. As one of its strengths, we directly assessed abdominal fat accumulation using CT scanning. This allowed the role of fat deposition in the development of metabolic syndrome and its components to be examined more closely. Secondly, the sample size of our study was sufficiently large (almost 7,000 subjects), and both sexes were included. Thirdly, we adjusted for alcohol consumption and physical activity, which may confound the association between abdominal fat accumulation and metabolic risk factors. However, the study was limited because of its cross-sectional design, and changes in the metabolic risk profile were not monitored.
In this study, we demonstrated a strong association between the combined effect of an increasing VFA and a decreasing adiponectin level with multiple risk factors of metabolic syndrome. Both the VFA and the adiponectin level were independently associated with the clustering of metabolic risk factors in a large Japanese population, which has a relatively low BMI compared with other ethnicities. The present findings have important implications for the prevention of metabolic syndrome. Further prospective studies are needed to assess the impact of the VFA and adiponectin level on the incidence of metabolic syndrome or cardiovascular diseases.
- 4World Health Organization. Measuring Obesity: Classification and Description of Anthropometric Data. Geneva: World Health Org., 1989 (Nutr UD, EUR/ICP/NUT 125).