INTRODUCTION

Osteoporosis is a prevalent systemic bone disease characterized by reduced bone mass and deteriorated microarchitecture, significantly increasing fracture risk, particularly in the elderly¹. As the global population ages, the incidence of osteoporosis is rising, resulting in millions of fractures each year and imposing substantial economic burdens on healthcare systems². While bone mineral density (BMD) is commonly used to diagnose and manage osteoporosis by assessing bone strength, it does not always correlate with fracture risk, as it may appear normal even when the risk remains high³. Therefore, BMD alone is insufficient for fully understanding bone health, particularly regarding bone microarchitecture. The trabecular bone score (TBS), derived from the analysis of dual-energy absorptiometry (DXA) images, provides a detailed assessment of bone microarchitecture, including trabecular distribution, connectivity, and structural integrity. TBS is calculated based on pixel variations in two-dimensional images, with dense trabecular structures exhibiting smaller and more numerous variations, while sparse structures show larger and fewer pixel differences⁴. Recent studies suggest that combining TBS with the Fracture Risk Assessment Tool (FRAX) enhances fracture risk prediction accuracy beyond that achieved with FRAX alone, offering a more comprehensive evaluation of fracture risk, especially in cases where BMD may not fully reflect bone strength⁵. Several studies have demonstrated a strong correlation between TBS and fracture risk, suggesting that TBS may be a more reliable predictor of fracture risk than BMD.

Smoking is a major risk factor for both osteoporosis and fractures⁶. Cotinine, a metabolite that remains in the body for an extended period, is recognized as a key biomarker for accurately determining tobacco smoke exposure⁷. Blood cotinine levels are more stable than those in urine, making serum cotinine a more dependable measure for evaluating tobacco exposure⁸. Cotinine not only reflects the level of tobacco exposure but is also linked to the negative effects of smoking on bone health, such as increased oxidative stress and the disruption of mesenchymal stem cell function, both critical for bone formation⁹. Despite the established link between smoking and bone damage, the connection between serum cotinine and TBS remains unclear.

The goal of this study is to examine the connection between serum cotinine levels and TBS in US participants, using data from the National Health and Nutrition Examination Survey (NHANES). Our hypothesis is that higher serum cotinine levels may be linked to lower TBS values.

METHODS

This research is a cross-sectional study utilizing data from the US NHANES database (http://www.cdc.gov/nchs/nhanes.htm), which reflects the health and nutritional status of the US population through a broad, representative survey¹⁰. The study protocol was reviewed and approved by the Ethics Review Board of the National Center for Health Statistics, with the most recent review in August 2022. All participants provided written informed consent. NHANES is a research program designed to assess the health and nutritional status of adults and children across different racial, ethnic, and socioeconomic groups in the United States. As TBS data were only available in the 2005–2008 NHANES cycles, the study included participants from the 2005–2006 and 2007–2008 periods. Initially, 20497 participants were considered, but after excluding those aged <20 years (n=9583), and those without TBS data (n=2859) or serum cotinine information (n=1094), the final analysis included 6961 individuals aged ≥20 years (Figure 1).

Figure 1

Flowchart for inclusion of participants, aged ≥20 years, with complete data on TBS and serum cotinine, NHANES 2005–2008 (N=6961)

RESULTS

Subgroup analysis

Table 4 presents a linear regression analysis of the specific relationship between serum cotinine levels and TBS. A significant association was found in females (β= -0.02; 95% CI: - 0.03 – -0.01). In the BMI-stratified subgroup, the interaction term had a p-value of 0.09. Although this did not meet the conventional significance threshold (p<0.05), it may still be considered marginally significant, indicating a potential interaction effect. Figures 2A–2D depict the associations between log-transformed serum cotinine and TBS across different populations, adjusting for sex, race, BMI, and PIR.

Table 4

Association between serum cotinine levels and trabecular bone score of participants, aged ≥20 years, with complete data on TBS and serum cotinine, stratified by sex, race, BMI, and PIR, NHANES 2005–2008 (N=6961)

	Model 1 β (95% CI)	Model 2 β (95% CI)	Model 3 β (95% CI)	p for interaction
Stratified by sex				0.017
Female	-0.008 (-0.019–0.002)	-0.035 (-0.046 – -0.024)	-0.021 (-0.033 – -0.008)
Male	0.024 (0.011–0.038)	-0.008 (-0.019–0.003)	-0.009 (-0.022–0.005)
Stratified by race				0.224
Black	0.008 (-0.016–0.033)	-0.007 (-0.028–0.014)	-0.003 (-0.024–0.018)
Other	-0.01 (-0.039–0.020)	-0.031 (-0.060 – -0.002)	-0.024 (-0.057–0.009)
Mexican American	0.002 (-0.019–0.022)	-0.015 (-0.034–0.004)	0.006 (-0.012–0.023)
White	0.01 (-0.001–0.022)	-0.029 (-0.040 – -0.019)	-0.012 (-0.025–0.001)
Stratified by BMI				0.09
<25	0.004 (-0.006–0.015)	-0.018 (-0.026 – -0.009)	-0.015 (-0.025 – -0.006)
25–29.9	0.003 (-0.014–0.019)	-0.03 (-0.045 – -0.015)	-0.024 (-0.037 – -0.011)
>29.9	-0.001 (-0.023–0.022)	-0.036 (-0.055 – -0.018)	-0.027 (-0.044 – -0.010)
Stratified by PIR				0.425
≤1	0.022 (-0.006–0.051)	0.002 (-0.017–0.021)	-0.014 (-0.031–0.004)
1.1–3.0	0.017 (0.003–0.031)	-0.021 (-0.035 – -0.006)	-0.016 (-0.030 – -0.002)
>3.0	0.005 (-0.009–0.020)	-0.023 (-0.036 – - 0.010)	-0.008 (-0.023–0.008)

[i] Model 1: no covariates were adjusted. Model 2: adjusted for sex, race, and age. Model 3: adjusted as in Model 2 plus PIR, BMI, total spine BMD, smoke, CRP, total protein, blood urea nitrogen, serum creatinine, serum uric acid, serum calcium, serum alkaline phosphatase, serum phosphorus. For abbreviations see Table 1.

Figure 2

Association between log2-transformed cotinine and the prevalence of trabecular bone score (TBS) in different subgroups among participants, aged ≥20 years, with complete data on TBS and serum cotinine, NHANES 2005–2008 (N=6961)

DISCUSSION

This large-scale cross-sectional study, which analyzed combined data from the 2005 to 2008 NHANES, included a total of 36463 participants. We found that serum cotinine levels were independently associated with TBS. After adjusting for all covariates, our analysis showed a significant negative relationship between serum cotinine levels and TBS among US adults. Thus, we conclude that smoking cessation or reducing exposure to secondhand smoke may help improve TBS, enhance bone health, and lower the risk of fractures.

To the best of our knowledge, there is a limited amount of research exploring the connection between serum cotinine and TBS. TBS is gaining attention as a novel approach for evaluating bone quality, as it provides insights into skeletal microarchitecture. In contrast to BMD, which primarily quantifies bone mineral content, the TBS offers a detailed assessment of bone microarchitecture, specifically the quality and arrangement of trabeculae. This microarchitectural information allows TBS to detect more nuanced impairments in bone quality due to smoking. Some studies have highlighted TBS’s ability to differentiate between individuals with similar lumbar spine BMD (lsBMD) values, serving as a valuable complement to lsBMD. Furthermore, evidence suggests that integrating the FRAX with TBS can significantly enhance the precision of fracture risk prediction^15,16. Research has shown that individuals who are actively or passively exposed to tobacco are at a higher risk of developing various health conditions, including reduced bone mass and an increased likelihood of osteoporotic fractures^17-19. Cotinine is often chosen as the primary biomarker for assessing tobacco exposure because it tends to have higher concentration levels and a longer elimination half-life^20,21. A cross-sectional study has suggested that reducing cigarette exposure and maintaining lower serum cotinine levels may promote bone health in adults, especially in women. This contrasts with the findings of our study regarding cotinine and BMD. A possible explanation lies in our use of TSBMD, as lsBMD data were unavailable in the NHANES 2007–2008; TSBMD spans multiple vertebral regions, potentially introducing regional variability that may have obscured the association between cotinine and BMD. Although many studies have demonstrated the connection between smoking and osteoporosis, the relationship between serum cotinine and TBS is still not well understood. In our study, we found a negative correlation between serum cotinine and TBS, which could be explained by several potential mechanisms.

Nicotine, the key addictive substance in cigarettes, interacts with nicotinic receptors found on osteoblasts²². Studies indicate that nicotine might impair the function of osteoblasts, thereby reducing bone matrix production, while also enhancing the activity of osteoclasts, which contributes to additional bone loss²³. Nicotine is unstable, with the vast majority of absorbed nicotine rapidly metabolized into cotinine. While nicotine has a short half-life in the bloodstream (2–4 hours), cotinine has a longer half-life (16–20 hours), making it a more reliable indicator of long-term and sustained nicotine exposure, rather than the immediate effects of a single smoking event^7,21. Smoking not only exerts a direct effect on bone health but also indirectly heightens the risk of osteoporosis through a variety of systemic mechanisms. Studies have shown that oxidative stress caused by smoking can trigger widespread inflammation, a chronic low-grade inflammatory state that is strongly linked to bone loss^24,25. In smokers, the levels of inflammatory cytokines such as TNF-α and IL-6 are elevated, which stimulates osteoclast activity and accelerates bone resorption^26,27. Vitamin D is essential for calcium absorption and bone mineralization, and its deficiency is directly linked to decreased bone density and the development of osteoporosis. Numerous studies have demonstrated that smokers tend to have lower serum vitamin D levels compared to non-smokers, suggesting that smoking may suppress the production of parathyroid hormone, calcifediol, and calcitriol^28,29. Furthermore, research by Kassi et al.³⁰ indicates that this deficiency is independent of age and sex, with smokers showing a 58% to 63% greater likelihood of vitamin D3 deficiency compared to non-smokers. Recently, there has been growing interest in the role of gut microbiota in osteoporosis, particularly regarding the composition of gut bacteria and its influence on bone health. Smoking has been found to disrupt gut microbiota balance and increase intestinal permeability, allowing harmful substances to enter the bloodstream and induce systemic inflammation^31,32. This imbalance may lead to reduced nutrient absorption, which indirectly compromises bone health. In our study, when stratified by sex, the association was significant only among females (β= -0.02; 95% CI: -0.03 – -0.01, p<0.05). This result suggests that smoking may exert a stronger negative effect on bone health in women, possibly due to the regulatory influence of hormones like estrogen on bone metabolism. Smoking also affects hormone levels, particularly estrogen and testosterone. Testosterone plays a role in promoting osteoblast proliferation by binding to androgen receptors on these cells, while estrogen helps suppress osteoclast activity^33,34. Although the effect of smoking on testosterone levels is still debated, its impact on women is clearer³⁵. In women, cotinine can inhibit aromatase activity and enhance the hepatic breakdown of estradiol into 2-methoxyestrone, resulting in lower circulating free estrogen levels. This reduction leads to decreased osteoblast activity and proliferation, along with increased osteoclast activity and bone resorption³⁶.

CONCLUSIONS

Our research identified an independent association between higher cotinine levels and lower TBS. This finding enhances our comprehension of the harmful impact of cigarette smoke on bone quality and underscores the critical role of smoking cessation in maintaining bone health. To substantiate these results, additional large-scale prospective studies are required.