Linking metabolic syndrome with low bone mass through insights from BMI and health behaviors

The objective of this study is to investigate the relationship between metabolic syndrome (MetS), and bone density in a 30- to 50-year-old Taiwanese population, and to explore the combined effects of BMI and health behaviors on this association. A total of 52,912 individuals aged 30–50 years from the Taiwan Biobank were included in this cross-sectional study. Bone density status was assessed using quantitative ultrasound (QUS). The joint effect was assessed by including an interaction term in the multi-logistic regression models to test the association between MetS, BMI, and bone density while controlling for potential confounders. MetS was associated with reduced bone density, with the risk of severe low bone density (SLBD) higher among BMI < 24 kg/m2 individuals with MetS (adjusted odds ratio [aOR] 1.5, 95% confidence interval [CI] 1.09–2.16), while the risk was not significant among BMI ≥ 24 kg/m2 individuals with MetS. Smoking, alcohol consumption, and lack of regular exercise among individuals with a BMI < 24 kg/m2 and MetS were associated with higher risk of severe low bone density (SLBD), the aORs (95%CI) were 2.9 (1.59–5.20), 2.1 (1.06–4.22), and 1.8 (1.24–2.54) respectively. Our study suggests that metabolic syndrome could increase the risk of severe low bone density, but this risk can be minimized through higher BMI, non-smoking, no alcohol consumption, and regular exercise. Conversely, smoking, alcohol consumption or lack of regular exercise may exacerbate the risk of severe low bone density. These findings highlight the importance of a multifactorial approach in managing bone healthcare.

Bone status.While DXA is considered the gold standard for measuring BMD, it has limitations such as radiation exposure, cost and limited availability for field studies.We used quantitative ultrasound (QUS) as a non-invasive, radiation-free alternative to measure bone status in the TWB cohort.The Lunar Achilles Insight ultrasonometer (GE Healthcare, Milwaukee, WI, USA) was used to take measurements at the right calcaneus 31 .QUS analyzes the transmission of sound waves through bone tissue to estimate bone status.The stiffness index (SI) was calculated from the BUA and SOS parameters, and the T-score was estimated based on the SI.The T-score corresponds to the number of standard deviations from the mean value of the population aged 20-39 years of the same sex.Correlations between bone parameters measured by ultrasonometry, BMD, and bone mineral content measured by DXA have been observed in the general population 32 .
In order to investigate the association between bone status and metabolic syndrome, we categorized participants' bone status according to their T-scores.As the World Health Organization defines osteopenia and osteoporosis only in populations over 50 years of age, we used the following classification for our study: T-scores between − 1.5 and − 2.5 were classified as mild low bone density (MLBD), while T-scores ≤ − 2.5 were classified as severe low bone density (SLBD).

Statistical analysis.
Descriptive statistics were used to summarize the demographic characteristics and health information of the study population.Continuous variables were presented as means ± standard deviations (SD), while categorical variables were presented as percentages.Differences in each variable across bone status were tested using either simple regression or logistic analysis.Multi-nominal logistic regression analysis was then performed to investigate the association between metabolic syndrome and bone status.Specifically, Model 1 was formulated to compare the odds values of MLBD and SLBD relative to the normal bone density group.Meanwhile, Model 2 aimed to compare the odds values of SLBD against MLBD.These analyses were adjusting for potential confounding factors including age, gender, educational level, marital status, vegetarianism, cigarette smoking, alcohol drinking, regular exercise and BMI.To examine the joint effect of MetS and overweight/ obesity on bone status, we created four categories based on the presence or absence of MetS and overweight/ obesity: (1) neither MetS nor overweight/obesity, (2) MetS only, (3) overweight/obesity only, and (4) both MetS and overweight/obesity.The interaction effect between MetS and overweight/obesity was tested by including an interaction term in the multi-nominal logistic regression model.The association further explored the joint effect with cigarette smoking, alcohol consumption and regular exercise, with all statistical analyses being conducted using STATA SE 17.0 (StataCorp LLC, College Station, TX, USA).A p-value < 0.05 was considered statistically significant.

Results
Demographic characteristics.A total of 52,912 individuals participated in the study, of whom 38.0% were male, and the mean age was 39.5 ± 5.6 years.The demographic characteristics of the study population according to bone density status are presented in Table 1.The proportion of low bone density was 22.8%, of which 19.5% were MLBD and 3.3% as SLBD.Low bone density was significantly associated with male gender, older age, lower education level, unmarried status, cigarette smoking, alcohol consumption, and lack of regular exercise.www.nature.com/scientificreports/Association between bone status and metabolic syndrome.Low bone density was associated with unfavorable results of metabolic syndrome components, including larger waist circumference, higher systolic and diastolic blood pressure, higher triglyceride and fasting glucose levels, and lower high-density lipoprotein cholesterol.This association was consistent across different strata of BMI (as shown in Fig. 1).Lower bone density was also associated with presence of self-reported disease, including DM, hypertension, dyslipidemia, cardiovascular disease and having a higher prevalence of MetS, with the results being consistent across different strata of BMI (Fig. 2).

Effect of BMI on the association between low bone density and metabolic syndrome.
To examine the effect of BMI on the association between low bone density and metabolic syndrome, the study population was further categorized based on the presence of overweight/obesity (BMI ≥ 24 kg/m 2 ) and MetS with adjusting for potential confounders, including age, gender, educational level, marital status, alcohol drinking, cigarette smoking, regular exercise, vegetarianism and BMI using multi-nominal logistic regression analysis (Table 2).The results showed that among individuals with a BMI < 24 kg/m 2 , those with MetS had a significantly higher risk of SLBD than those without MetS (adjusted odds ratio [aOR] 1.5, 95% confidence interval [CI] 1.09-2.16).On the other hand, among individuals without MetS, those who have a BMI ≥ 24 kg/m 2 had a significantly lower risk of SLBD compared to those with a BMI < 24 kg/m 2 (aOR 0.8, 95% CI 0.67-0.92).For individuals with a BMI ≥ 24 kg/m 2 and MetS, the association with the risk of having SLBD was insignificant (aOR 1.0, 95% CI 0.79-1.21)compared to those with a BMI < 24 kg/m 2 and without MetS.In the Model 2 analysis, individuals with a BMI ≥ 24 kg/m 2 also had a significant lower risk of SLBD, irrespective of their MetS status.

Effect of health behaviors on the association between BMI, low bone density and metabolic syndrome.
Table 3 presents the effect of cigarette smoking on the association between MetS and SLBD stratified by obesity status.The analysis was adjusting for potential confounders, including age, gender, educational level, marital status, alcohol drinking, regular exercise, vegetarianism and BMI.In general, the risks of SLBD were either insignificant or lower among non-smokers, irrespective of their BMI and MetS status.Among smokers, individuals with a BMI < 24 kg/m 2 , irrespective of MetS status, and those with a BMI ≥ 24 kg/m 2 in conjunction with MetS, had higher risks of SLBD compared to the reference group.Table 4 presents the effect of alcohol consumption on the association between MetS and SLBD stratified by obesity status.The analysis adjusted for potential confounders, including age, gender, educational level, marital status, cigarette smoking, regular exercise, vegetarianism and BMI.Among non-drinkers, individuals with a BMI ≥ 24 kg/m 2 and without MetS had lower risk of SLBD compared the reference group.Individuals who were drinkers with a BMI < 24 kg/m 2 and with MetS had significantly higher risk of SLBD compared the reference group (aOR 2.1, 95% CI 1.06-4.22),while the association was insignificant in non-drinkers.Table 5 shows the effect of regular exercise on the association between MetS and SLBD stratified by obesity status.The analysis adjusted for potential confounders, including age, gender, educational level, marital status, cigarette smoking, alcohol drinking, vegetarianism and BMI.Individuals who had regular exercise and without MetS, regardless of BMI status, had lower risk of SLBD compared to those who did not have regular exercise (aOR 0.7, 95% CI 0.60-0.83and aOR 0.4, 95% CI 0.33-0.54,for BMI < 24 and ≥ 24 kg/m 2 respectively).Individuals who had no regular exercise, with BMI < 24 kg/m 2 and with MetS had higher risk of having SLBD compared to BMI < 24 kg/m 2 and without MetS (aOR 1.8, 95% CI 1.24-2.54),while the association was insignificant in the regular exercise group (aOR 0.4, 95% CI 0.12-1.23).To better understand the relationship between SLBD and MetS, and the effect of BMI, cigarette smoking, alcohol drinking and exercise on the association, the aORs for SLBD found in Tables 2, 3, 4, 5 is summarized into Table 6.

Discussion
Our results show that metabolic syndrome was associated with reduced low bone density, with the association being consistent across normal, overweight and obesity groups among people aged 30-50 from a TWB cohort.Subsequent stratification of metabolic syndrome by overweight/obesity (using a BMI cut-off of 24 kg/m 2 ) and controlling for the potential confounding factors indicated that normal weight individuals with metabolic syndrome were at an increased risk of lower bone density.Conversely, the risk of lower bone density was attenuated in overweight/obese individuals with metabolic syndrome.Additionally, overweight/obese individuals without metabolic syndrome were found to be at a lower risk of SLBD compared to normal weight individuals.www.nature.com/scientificreports/ The results suggest that metabolic syndrome, by itself, could increase the risk of bone loss, and an increased BMI could serve as a protective function in reducing bone loss.Thus, the risk of SLBD in individuals with overweight/obesity and metabolic syndrome is mitigated through the effect of mechanical loading from body weight.A study investigating the association between metabolic syndrome and BMD among adolescents found that metabolic syndrome was associated with a lower concentration of bone biomarkers for osteocalcin, bone alkaline phosphatase and carboxy-terminal telopeptide, indicating reduced bone formation and resorption 33 .Hwang D.K. et al. conducted an analysis of the association between BMD and metabolic syndrome in Korean women aged 18 years and above, with the results revealing that an increased number of metabolic syndrome components was linked to low vertebral BMD 34 .Some studies have previously reported a negative association  www.nature.com/scientificreports/ between MetS and osteoporosis, with this association being diminished after controlling for body mass index (BMI) and other potential confounding factors 26,35 , indicating that the inverse relationship between MetS and osteoporosis might be largely explained by the increased BMI levels observed in individuals with MetS.
The findings of our study provide further evidence supporting this relationship, although in a case-control study that investigated the association between MetS and osteoporosis in adults aged 50 years and above, the findings were inconsistent with our study, demonstrating that MetS was associated with a low occurrence of osteoporosis, but was associated with a high occurrence of osteoporosis among obese men and obese post-menopausal women 24 .One possible explanation for the different findings is that our study population was comprised of younger, pre-menopausal women, suggesting the possibility of a generational difference in the association.Moreover, it is plausible that the duration of obesity and MetS in older populations might have had an impact on bone health; additionally, most studies have examined the effect of MetS and obesity on osteoporosis independently, while our study examined the joint effect of MetS and obesity on severe low bone density, providing an insight into the complex relationship between MetS, obesity and osteoporosis.
Cigarette smoking is recognized as an independent risk factor for the reduction in bone mass and bone length, development of osteoporosis, and increased risk of fractures [36][37][38] .The effects of cigarette smoking on bone Table 5. Adjusted odds ratios of bone status associated with BMI strata with or without metabolic syndrome and exercise.Model 1: Used 'BMI < 24 with no MetS and no regular exercise' as reference group.The aORs were adjusted for age, gender, educational level, marital status, cigarette smoking, alcohol drinking, vegetarian status and BMI.Model 2: Used 'BMI < 24 with no MetS and no regular exercise' as reference group, and compared the odds of SLBD relative to MLBD.The aORs were adjusted for age, gender, educational level, marital status, cigarette smoking, alcohol drinking, vegetarian status and BMI.BMI body mass index, MetS metabolic syndrome, MLBD mild low bone density, denotes T score between − 1.5 and − 2.   39 .
Our study demonstrated that smokers with normal weight and without MetS, or with MetS regardless of their BMI status, had significantly higher risk of SLBD, when compared to the reference group with normal weight, without metabolic syndrome, and non-smokers.Specifically, individuals with normal weight and metabolic syndrome who smoke had the highest risk of SLBD.These findings suggest that the deleterious effects of smoking on bone health might be more pronounced than the beneficial effects of mechanical loading from body weight; furthermore, smokers are more likely to have lower body weight, which could further exacerbate the negative impact on bone health.
Chronic alcohol consumption, like cigarette smoking, has been found to have both direct and indirect effects on bone metabolism 40 .Animal models have shown that alcohol consumption could disrupt calcium homeostasis through effects on the parathyroid hormone-vitamin D-axis 41 and growth hormone-insulin-like growth hormone signaling 42 , and inhibit bone formation and fracture repair through several molecular targets 43,44 .The consequences of chronic alcohol consumption on bone include a loss of bone mineral density, impaired bone quality, and an increased risk of osteoporosis 45 and osteoporotic hip fracture, with a dose-response relationship observed 46 .However, the impact of lower doses of alcohol on bone health remains uncertain.Interestingly, some studies suggest that light drinking might have a beneficial effect on bone mineral density 47 , as it has been observed that BMD is even higher in light drinkers compared to abstainers 46 .Our study demonstrated that alcohol consumption was not significantly associated with SLBD, except in the case of normal-weight individuals with metabolic syndrome, who had significantly higher risk of SLBD.However, the lack of consideration for the amount of alcohol consumed by the participants complicates the interpretation of the results, and accordingly, the association between alcohol consumption and SLBD could not be fully elucidated.
Past studies have reported very consistent results on the beneficial effects of exercise on BMD of the lumbar spine and the femur in both menopausal and more elderly women.Numerous publications have linked physical exercise, bone metabolism markers, and bone mineral density 48 .Our study found that the risk of SLBD was elevated in individuals with normal weight and metabolic syndrome who did not have regular exercise, but was not significant in those who had regular exercise.Additionally, the risk of SLBD was lower in individuals without metabolic syndrome who engaged in regular exercise regardless of their BMI status when compared to the reference group with normal weight, without metabolic syndrome, and non-regular exercisers.Although we did not differentiate the type or intensity of the exercise, these results suggest that, in general, the beneficial effect of regular exercise on bone density could outweigh the negative effect of low BMI and metabolic syndrome.
Our study has several strengths and limitations.Firstly, we utilized a large population-based cohort, enabling us to explore the complex relationship between obesity, metabolic syndrome, health behaviors and bone density.Secondly, we examined the joint effect of metabolic syndrome and obesity, as well as the combined effects of cigarette smoking, alcohol consumption and exercise on bone density, providing additional information on understanding how BMI and health behaviors modify the relationship in early life.However, our study was cross-sectional, which means that a causal relationship between the factors we examined cannot be established, while additionally, our study relied on self-reported data on lifestyle behaviors, which might be subject to reporting bias, with any such bias being non-differentiated, leading to potentially underestimated results.Finally, our study did not consider quantitative information regarding health behaviors such as the amount of smoking and alcohol consumption or the intensity of exercise, which might have important implications for bone health.

Conclusion
Our study suggests that metabolic syndrome could increase the risk of severe low bone density, but this risk can be minimized through higher BMI, non-smoking, no alcohol consumption, and regular exercise.Conversely, smoking, alcohol consumption or lack of regular exercise could exacerbate the risk of severe low bone density.These findings highlight the importance of a multifactorial approach in managing bone healthcare.

Figure 1 .
Figure 1.Relationship between components of MetS and bone density status in different BMI strata.The Y-axis represents the value of each metabolic syndrome component, the unit of each factor is indicated in the subtitle of the respective graph.BMI body mass index, MetS metabolic syndrome, MLBD mild low bone density, SLBD severe low bone density.*P trend < 0.05.

Figure 2 .
Figure 2. The association of metabolic syndrome and bone density status across different BMI strata.BMI body mass index, MetS metabolic syndrome, MLBD mild low bone density, SLBD severe low bone density.*p < 0.05 for comparison to normal bone density group.

Table 1 .
Demographic characteristics of the study population in relation to bone status.Participants with T-score ≥ − 1, − 2.5 ~ − 1, and ≤ − 2.5 were defined as normal, MLBD, and SLBD, respectively.MLBD moderate low bone density, SLBD severe low bone density, BMI body mass index, MetS metabolic syndrome.*p < 0.05. 1 Diff.1 and Diff.2 denote the difference in mean or proportion for MLBD vs. normal and SLBD vs. normal, respectively. 2p trend values for increasing or decreasing difference across bone status.

Table 2 .
Adjusted odds ratios of bone status associated with BMI strata with or without metabolic syndrome.Abbreviations: BMI, body mass index; MetS, metabolic syndrome; MLBD, mild low bone density, denotes T score between − 1.5 and − 2.5; SLBD, severe low bone density, denotes T-scores ≤ − 2.5; Ref., reference; *p < 0.05.Model 1: Used 'BMI < 24 with no MetS' as reference group.The aORs were adjusted for age, gender, educational level, marital status, alcohol drinking, cigarette smoking, regular exercise, being a vegetarian and BMI.Model 2: Used 'BMI < 24 with no MetS' as reference group, and compared the odds of SLBD relative to MLBD.The aORs were adjusted for age, gender, educational level, marital status, alcohol drinking, cigarette smoking, regular exercise, being a vegetarian and BMI.

Table 3 .
Adjusted odds ratios of bone status associated with BMI strata with or without metabolic syndrome and cigarette smoking.Model 1: Used 'BMI < 24 with no MetS in non-smokers' as reference group.The aORs were adjusted for age, gender, educational level, marital status, alcohol drinking, regular exercise, being a vegetarian and BMI.Model 2: Used 'BMI < 24 with no MetS in non-smokers' as reference group.and compared the odds of SLBD relative to MLBD.The aORs were adjusted for age, gender, educational level, marital status, alcohol drinking, regular exercise, being a vegetarian and BMI.BMI body mass index, MetS metabolic syndrome, MLBD mild low bone density, denotes T score between − 1.5 and − 2.5, SLBD severe low bone density, denotes T-scores ≤ − 2.5., Ref. reference.*p < 0.05.

Table 4 .
Adjusted odds ratios of bone status associated with BMI strata with or without metabolic syndrome and alcohol consumption.Model 1: Used 'BMI < 24 with no MetS in non-alcohol drinkers' as reference group.The aORs were adjusted for age, gender, educational level, marital status, cigarette smoking, regular exercise, vegetarian status and BMI.Model 2: Used 'BMI < 24 with no MetS in non-alcohol drinkers' as reference group, and compared the odds of SLBD relative to MLBD.The aORs were adjusted for age, gender, educational level, marital status, cigarette smoking, regular exercise, vegetarian status and BMI.BMI body mass index, MetS metabolic syndrome, MLBD mild low bone density, denotes T score between − 1.5 and − 2.5, SLBD severe low bone density, denotes T-scores ≤ − 2.5, Ref. reference.*p < 0.05.

Table 6 .
Summary of adjusted odds ratios for SLBD associated with metabolic syndrome in condition of BMI status and health behaviors.BMI body mass index, MetS metabolic syndrome, Ref. reference, NS nonsignificance.metabolismmight involve both indirect and direct mechanisms.Indirect mechanisms could include decreases in body weight, change in the parathyroid hormone-Vitamin D-Axis, gonadal hormones and increased oxidative stress, while other studies have demonstrated that cigarette smoking directly impacts bone tissue by binding to various receptors, such as nicotinic acetylcholine receptors and androgen receptors in osteoblasts, as well as aryl hydrocarbon receptors in osteoclasts