Osteoporosis International

, Volume 21, Issue 3, pp 425–431

The relationship between low bone mass and metabolic syndrome in Korean women

Authors

  • D.-K. Hwang
    • Family MedicineEulji University Hospital
    • Family MedicineEulji University School of Medicine
Original Article

DOI: 10.1007/s00198-009-0990-2

Cite this article as:
Hwang, D. & Choi, H. Osteoporos Int (2010) 21: 425. doi:10.1007/s00198-009-0990-2

Abstract

Summary

We examined the relationship between low bond mass and metabolic syndrome in 2,475 Korean women. After adjustment for all covariates, mean vertebral BMD was significantly lower in women with metabolic syndrome. Moreover, age and weight adjusted vertebral BMD was significantly decreased with additional components of the metabolic syndrome.

Introduction

Obesity-induced chronic inflammation is a key component in the pathogenesis of insulin resistance and metabolic syndrome. It has been suggested that proinflammatory cytokines and low-grade systemic inflammation activate bone resorption and may lead to reduced bone mineral density (BMD). The objective of this study was to determine the relationship between low bone mass and metabolic syndrome in Korean women.

Methods

This is a cross-sectional study of 2,548 women aged 18 years and over who had visited the Health Promotion Center. Physical examination and laboratory tests were performed. Vertebral BMD was measured using dual-energy X-ray absorptiometry. Metabolic syndrome was defined by National Cholesterol Education Program—Adult Treatment Panel III criteria.

Results

Among 2,475 women, 511 (21.0%) women had metabolic syndrome. Women with abdominal obesity or hypertriglyceridemia had significantly lower vertebral BMD than women without respective components after adjustment for age, weight, and height. After adjustment for all covariates, mean vertebral BMD was significantly lower in women with metabolic syndrome (p = 0.031). Moreover, age- and weight-adjusted vertebral BMD were significantly decreased with additional components of the metabolic syndrome (p = 0.004).

Conclusions

These findings suggest that metabolic syndrome might be another risk factor for osteoporosis and related fractures.

Keywords

Body mass index (BMI)Bone mineral density (BMD)Metabolic syndromeObesityOsteoporosisWomen

Introduction

Osteoporosis is a skeletal disease characterized by low bone mass and microarchitectural deterioration of bone tissue with a consequent increase in bone fragility and susceptibility to fractures [1]. Osteoporotic fractures are associated with increased mortality and morbidity. Survival rates 5 years after hip and vertebral fractures were found in Rochester, MN, to be 80% of those expected for men and women of similar age without fractures [2]. In the USA, 6.7% of survivors of all types of osteoporotic fractures become dependent in basic activities of daily life and 7.8% require long-term nursing home care [3]. Therefore, osteoporosis is a disease that has a major impact on public health with an increasingly elderly population.

Metabolic syndrome, another worldwide health problem, is a cluster of risk factors that are associated with cardiovascular diseases and type-2 diabetes [4]. The major characteristics of metabolic syndrome include abdominal obesity, lipid abnormality (i.e., elevated levels of triglycerides and low levels of high-density lipoprotein cholesterol (HDL-C)), hypertension, and hyperglycemia. The 2001 National Cholesterol Education Program—Adult Treatment Panel III (NCEP-ATP III) definition requires the presence of three or more of the components mentioned above [5]. The overall prevalence of metabolic syndrome in the USA is 22.6% in women and increases with age [6]. Kim and colleagues [7] reported that the prevalence of metabolic syndrome among women was 27.8% in Korea. Moreover, its rise according to age was steeper in women, and one of two women acquired metabolic syndrome. Abdominal obesity is not only a component of metabolic syndrome but also a cause of metabolic syndrome inducing insulin resistance. In previous observational studies, high body mass index (BMI) or obesity is associated with high bone mass and protects against excessive bone loss in aging. In a US population-based study, femoral neck bone mineral density (BMD) was associated with abdominal obesity, and femoral neck BMD increased along with additional components of metabolic syndrome [8]. Another prospective study demonstrated that metabolic syndrome reduced the risk of non-vertebral fractures [9]. However, osteoporosis is common in various inflammatory conditions, and it has been suggested that inflammatory cytokines and low-grade systemic inflammation activate bone resorption and may lead to reduced BMD and non-traumatic fractures [10, 11]. A recent study demonstrated that fat mass was negatively associated with bone mass after mechanical loading effects due to total body weight were controlled [12]. In another study, incidence of osteoporotic non-vertebral fractures was higher in subjects with metabolic syndrome [12].

Thus, the association between metabolic syndrome and BMD has been researched, but the results are inconclusive. Moreover, there were no data on Korean subjects. Therefore, we conducted a cross-sectional study to examine the relationship between metabolic syndrome and vertebral BMD among Korean women.

Methods

Study population

We prepared the data of 2,548 women aged 18 years and over who had visited the Health Promotion Center at Eulji University Hospital from January 2006 and December 2007. Seventy-three women with a history of the following conditions were excluded from the study: those taking medications for osteoporosis; those receiving hormone therapy or other medications affecting BMD, such as glucocorticoid; those with conditions that would be apparent causes for abnormal bone mass, such as malignancy, thyroid disease, or abnormal thyroid function test; and those with acute infectious diseases or chronic inflammatory diseases. A total of 2,475 women were analyzed. Menopause was defined as the time when there is no menstrual period for 12 consecutive months or levels of follicle-stimulating hormone (FSH) were ≥40 mIU/mL. During subjects' Health Promotion Center visits, medical history and life-style-related data (alcohol consumption, smoking, and exercise status) were collected via a questionnaire and history taking. Current medication use was verified by an examination of prescriptions. Alcohol consumption was measured according to the amount of alcohol more than one bottle of soju (a kind of distilled spirits containing 56.8 g of pure alcohol) consumed per week. Regular exercise was defined as 30 min or more at a time, three times per week regularly. Height, weight, and waist circumference were measured in subjects while they were wearing a light robe and no shoes. Waists were measured using WHO waist circumference measurement guidelines [13]. BMI was calculated as body weight (kg) divided by height squared (m2). Blood pressure was measured with an automatic instrument (BP-8800C, Colin Electronics, Komaki, Japan) in a sitting position on at least two different occasions after 10 min of rest. The study was approved by the Institutional Review Board of the Eulji University Hospital, Daejon, Korea.

Definition of metabolic syndrome

Metabolic syndrome was defined by NCEP-ATP III criteria [5]. All components of metabolic syndrome were defined by NCEP-ATP III criteria except abdominal obesity and hyperglycemia. Abdominal obesity and hyperglycemia were redefined on the basis of Asia-Pacific criteria [14] and American Diabetes Association 2003 guidelines [15], respectively. Subjects were classified as having metabolic syndrome if any three of the following five characteristics were present: abdominal obesity (waist circumference greater than 85 cm), hyperglycemia (fasting glucose ≥ 100 mg/dL (5.6 mmol/L) and/or medication), hypertriglyceridemia (fasting triglyceride ≥ 150 mg/dL (1.7 mmol/L)), low high-density lipoproteinemia (HDL-C < 50 mg/dL (1.29 mmol/L), and hypertension (>130/85 mm Hg and/or medication)).

Laboratory measurements

During each subject's visit, morning blood samples were obtained after a requested 12-h fast. All samples were analyzed within 3 h after the sampling. Blood chemistry, such as serum glucose, lipid profile including low-density lipoprotein cholesterol (LDL-C) was measured by an enzymatic technique using an ADVIA 1650 analyzer (Bayer Diagnostics, Terrytown, NY, USA). We measured high-sensitive C-reactive protein as a marker of systemic low-grade inflammation, which was measured by turbidmetric immunoassay using a Hitachi 7170 S (Hitachi Hi-Tech, Tokyo, Japan). Additionally, insulin was measured by chemiluminescence using Advia Centaur XP (Siemens Healthcare Diagnostics, Los Angeles, CA, USA) and, as a maker of insulin resistance, homeostatic model assessment—insulin resistance (HOMA-IR) was calculated as \( \left[ {\left\{ {{\text{glucose }}\left( {{\text{mg}}/{\text{dL}}} \right)/18.1} \right\} \times {\text{ insulin}}} \right]/22.5 \).

Measurement of bone mineral density

Lumbar spine BMD was measured during the visit using dual-energy X-ray absorptiometry using a QDR-1000 (Hologic, Bedford, MA, USA), calibrated daily using a standard phantom provided by the manufacturer. Measurements were maintained within the manufacturers' precision standard. Lumbar spine BMD was the mean of lumbar vertebrae 1–4.

Data analyses

Values are expressed as number (%) or mean ± SD, unless otherwise noted. Statistical comparisons were performed using Student's t test to assess general characteristics between women with or without metabolic syndrome. Analysis of covariance (ANCOVA) was used to compare vertebral BMD between women with and without each component of metabolic syndrome, the vertebral BMD of women with and without metabolic syndrome, and vertebral BMD according to the number of components of metabolic syndrome after adjusting for known BMD covariates including age, weight, height, alcohol consumption, smoking status, exercise status, and menopause. Statistical significance was defined as p < 0.05. All statistical analyses were performed using SPSS (SPSS, Chicago, IL, USA, SPSS Categories® 11.0).

Results

Among 2,475 women, 511 (21.0%) women had metabolic syndrome. The mean age of the women with and without metabolic syndrome was 56.2 and 45.3 years, respectively (p < 0.001). The age range was 21–94 years old. Table 1 shows the general characteristics of the women with or without metabolic syndrome. There were no significant differences in alcohol consumption and smoking status except regular exercise between the two groups. Compared to women without metabolic syndrome, women with metabolic syndrome had significantly higher levels of BMI; waist circumference; lipid profiles, except for HDL-C; fasting glucose; white blood cell counts; C-reactive protein; and insulin (p < 0.001). HOMA-IR were significantly higher in women with metabolic syndrome (p < 0.001). Women with metabolic syndrome had a higher prevalence of diabetes and menopause.
Table 1

General characteristics of subject by metabolic syndrome status

 

Metabolic syndrome

Without (n = 1,964)

With (n = 511)

p value

Mean ± SD

Mean ± SD

Age (years)

45.3 ± 11.1

56.3 ± 9.7

0.000

BMI (kg/m2)

22.6 ± 2.9

25.9 ± 2.9

0.000

Height (cm)

157.1 ± 5.6

154.3 ± 5.5

0.000

Waist (cm)

75.4 ± 8.1

86.5 ± 6.8

0.000

Systolic blood pressure (mm Hg)

120.9 ± 16.0

140.7 ± 16.2

0.000

Diastolic blood pressure (mm Hg)

69.8 ± 10.1

79.5 ± 9.8

0.000

Fasting glucose (mg/dL)

87.4 ± 12.4

103.7 ± 29.8

0.000

Total cholesterol (mg/dL)

187.1 ± 33.9

202.9 ± 40.0

0.000

Triglyceride (mg/dL)

98.2 ± 46.6

197.5 ± 92.3

0.000

HDL-cholesterol (mg/dL)

62.2 ± 13.1

49.9 ± 10.1

0.000

LDL-cholesterol (mg/dL)

103.8 ± 28.7

115.1 ± 33.2

0.000

WBC count (×103/mm3)

6.0 ± 1.6

6.9 ± 1.8

0.000

CRP (mg/dL)

0.14 ± 0.36

0.25 ± 0.36

0.000

Insulin (mU/L)

5.68 ± 3.19

8.29 ± 4.49

0.001

HOMA-IR

1.24 ± 0.78

2.15 ± 1.52

0.000

  

N (%)

N (%)

Smoking N (%)

59 (3)

10 (2)

0.190

Alcohol drinking N (%)

98 (5)

25 (5)

0.936

Regular exercise N (%)

1,080 (55)

240 (47)

0.002

Hypertension N (%)

235 (12)

66 (13)

0.418

Diabetes N (%)

59 (3)

35 (7)

0.000

Menopause N (%)

707 (36)

393 (77)

0.000

p values are calculated by Student's t test.

LDL low-density lipoprotein, CRP C-reactive protein, HOMA-IR homeostatic model assessment—insulin resistance

Table 2 shows the vertebral BMD of women with or without each component of metabolic syndrome independent of metabolic syndrome. Women with abdominal obesity or hypertriglyceridemia had significantly lower vertebral BMD than women without the respective associated components after adjustment for age, weight, and height. However, there was no difference in vertebral BMD between women with low high-density lipoproteinemia, hypertension, or hyperglycemia and women without such a component.
Table 2

Comparison of vertebral BMD (g/cm2) of women with or without each component of metabolic syndrome

 

Without

With

p value

Abdominal obesity

0.922 ± 0.004 (n = 1,487)

0.894 ± 0.005 (n = 988)

0.000

Hypertriglyceridemia

0.916 ± 0.003 (n = 1,908)

0.895 ± 0.006 (n = 567)

0.002

Low high-density lipoproteinemia

0.913 ± 0.003 (n = 1,884)

0.903 ± 0.006 (n = 591)

0.129

Hypertension

0.915 ± 0.004 (n = 1,515)

0.905 ± 0.005 (n = 960)

0.094

Hyperglycemia

0.911 ± 0.003 (n = 2,211)

0.912 ± 0.008 (n = 264)

0.947

Values are expressed as least square mean ± SE. p values are calculated by analysis of covariance. All covariates: age, weight, height

In Table 3, unadjusted mean vertebral BMD of women with and without metabolic syndrome were 0.858 ± 0.007 and 0.925 ± 0.004 g/cm2, respectively (p < 0.001). Age-, weight-, and height-adjusted vertebral BMD of women with and without metabolic syndrome were 0.895 ± 0.007 and 0.915 ± 0.003 g/cm2, respectively (p = 0.009).
Table 3

Vertebral BMD (g/cm2) of women with or without the metabolic syndrome

 

Metabolic syndrome

Without (n = 1,964)

With (n = 511)

p value

Unadjusted

0.925 ± 0.004

0.858 ± 0.007

0.000

Age, height, weight adjusted

0.915 ± 0.003

0.895 ± 0.007

0.009

All covariates adjusted

0.914 ± 0.003

0.898 ± 0.007

0.031

Values are expressed as least square mean ± SE. p value is calculated by analysis of covariance (ANCOVA). All covariates: age, weight, height, alcohol consumption, exercise, menopause

After further adjustment for age, weight, height, alcohol consumption, smoking, exercise, and menopause status, mean vertebral BMD was consistently lower in women with metabolic syndrome (p = 0.031).

Figure 1 shows vertebral BMD according to the number of metabolic syndrome components. Age-, weight-, and height-adjusted vertebral BMD were significantly decreased with additional components of the metabolic syndrome (p = 0.004). The inclusion of other covariates (alcohol consumption, smoking, exercise, and menopause) did not change the result (p = 0.029) (data not shown).
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-009-0990-2/MediaObjects/198_2009_990_Fig1_HTML.gif
Fig. 1

Vertebral BMD (g/cm2) according to the number of the components of metabolic syndrome. p value is calculated by analysis of covariance (ANCOVA). Vertebral BMD levels are adjusted for age, weight, and height

C-reactive protein levels were significantly higher in women with metabolic syndrome after adjusting for all covariates (0.151 ± 0.010 vs 0.221 ± 0.022 mg/dL, respectively, p = 0.005). Additionally, adjusted C-reactive protein levels increased with more components: C-reactive protein = 0.12 mg/dL for no component, 0.16 mg/dL for one component, 0.17 mg/dL for two components, 0.24 mg/dL for three components, 0.21 mg/dL for four components, and 0.32 mg/dL for five components (p = 0.028).

Discussion

The prevalence of metabolic syndrome or osteoporosis increases with aging in women. Poor quality of life and increased morbidity and mortality are common features of these conditions. Metabolic syndrome is a cluster of conditions inducing cardiovascular disease or type-2 diabetes and a known association with low-grade inflammation [16]. Abdominal obesity is not only a key component of metabolic syndrome but also a cause of metabolic syndrome inducing insulin resistance. Many studies have already demonstrated the link between obesity and inflammation or insulin resistance [17, 18]. The plasma levels of various inflammatory mediators (IL-6, tumor necrosis factor-alpha, and C-reactive protein) are elevated in obesity. Also, insulin resistance states are associated with endothelial dysfunction and increased cardiovascular risk.

C-reactive protein is a sensitive systemic marker of inflammation and tissue damage [19]. Plasma C-reactive protein is only produced by hepatocytes, predominantly under transcriptional control by the cytokine IL-6, although other sites of local C-reactive protein synthesis and possible secretion have been suggested. Raised C-reactive protein levels are associated with many features of insulin resistance or metabolic syndrome [20], up to and including frank diabetes mellitus [21]. This may reflect, in part, the fact that adipocytes are the source of a substantial portion of IL-6 production [22].

On the other side of the equation, inflammatory cytokines up-regulate receptor activator of nuclear factor-κB ligand, leading to increased bone resorption and reduced BMD [23]. Some have suggested that an elevated C-reactive protein is associated with osteoporosis and non-traumatic fractures [10, 11].

So far, the associations of metabolic syndrome and low bone mass or osteoporotic fractures have not been extensively studied, and results are not conclusive. Recently, in a US population-based study, subjects with metabolic syndrome had an increased femoral neck BMD compared with controls without the syndrome. In the same study, initially unadjusted femoral neck BMD was reduced among persons with metabolic syndrome, but after adjustment for age, gender, and other covariates, it was higher in subjects with metabolic syndrome than in control groups. Moreover, in multivariate linear regression models for each component of metabolic syndrome, femoral neck BMD was significantly higher in subgroups of people with abdominal obesity and diabetes [8]. In another population-based prospective study, metabolic syndrome reduced risks of non-vertebral fractures. Also, reduced non-vertebral fracture risk was associated with high blood pressure in men and increased body mass in women. Additionally, lower non-fasting serum HDL-C protected against fractures in women and obese men [9]. Kinjo and colleagues [8] suggested that higher BMD with metabolic syndrome is largely determined by abdominal obesity, and protective effects of fat mass may promote bone formation via high mechanical loading, high circulating insulin levels, and factors that are co-secreted with insulin. On the other hand, a Rancho Bernardo Study analysis showed that the incidence of osteoporotic non-vertebral fractures was higher in subjects with metabolic syndrome. Also, after adjusting for BMI, a positive association between metabolic syndrome and femoral neck BMD was reversed, such that metabolic syndrome was associated with lower BMD [24]. In our study, vertebral BMD of women with metabolic syndrome was significantly lower than that of those without metabolic syndrome.

These two opposite results may be due to differences in subjects (race, age, gender, and co-morbid diseases status), adjusting covariates (especially menopause in women), employed sample (non-fasting serum), and methods of analysis. In this study, insulin levels and HOMA-IR were not correlated with vertebral BMD in spite of higher levels in women with metabolic syndrome (data not shown). Moreover, regardless of metabolic syndrome, women with abdominal obesity showed lower adjusted vertebral BMD than women without abdominal obesity.

Another study showed that the correlation between fat mass or percent of fat mass and bone mass was negative after adjustment for the mechanical loading effect of body weight [25]. In our study, vertebral BMD did not correlate with BMI, but it did correlate with weight and waist circumference (r = 0.197, r = −0.119, and p = 0.000, respectively). In the same way, we evaluated the relationship between metabolic syndrome and low bone mass by adjusting for the mechanical loading effects of body weight on bone mass.

Obesity and type-2 diabetes have both been known to be associated with increased BMD. Obesity or high BMI is known to be a protecting factor against excessive bone loss in aging [2629]. In a previous study, Barrett-Connor and colleagues reported that hyperinsulinemia may be responsible for part of the association of both diabetes and obesity with BMD in women [30]. However, because subjects with high insulin resistance showed more inflammation than subjects with low insulin resistance state [31, 32], more progressed inflammation in diabetes may eventually result in reduced BMD.

Previous studies suggest that osteoporosis is linked to inflammation [10, 33]; it is not yet clear whether higher C-reactive protein levels are associated with reduced bone mass. In this study, despite C-reactive protein levels being significantly higher in women with metabolic syndrome after adjusting for all covariates (p = 0.005), they were not correlated with vertebral BMD. These findings may be due to the counteracting of a negative effect of low-grade inflammation associated with metabolic syndrome and the protective effect of adiposity or diabetes.

Our study has several limitations. First, because our data are cross-sectional, we cannot assess causality between metabolic syndrome and low BMD. Second, because subjects were selected from women who visited a health promotion center, a selection bias may exist. Also, these results cannot be generalized to men or other population groups. Third, because we only measured vertebral BMD, we do not know the association between metabolic syndrome and BMD at other sites. Fourth, we did not investigate calcium intake or serum 25(OH)D3 levels, which are other factors affecting BMD.

In conclusion, we found that women with metabolic syndrome had lower vertebral BMD after adjusting for age, weight, height, alcohol consumption, smoking, exercise, and menopause. Also, regardless of metabolic syndrome, age-, weight-, and height-adjusted vertebral BMD were significantly lower with additional components of metabolic syndrome. These findings thus suggest that metabolic syndrome might be another risk factor for osteoporosis and related fractures.

Acknowledgements

This work was supported in part by The Korean Society of Menopause Grant.

Conflicts of interest

None.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2009