DOI:http://doi.org/10.65281/709817
Shan Su1, Yiding Zhao2 , Dece Kong2, Tieyi Yang2※
1Graduate School of Ningxia Medical University, Yinchuan 750004, Ningxia Hui Autonomous Region, China
2Department of Orthopedics, Pudong New Area Gongli Hospital, Shanghai 200135, China
Corresponding Author:
Tieyi Yang
Department of Orthopedics, Gongli Hospital, 219 Miaopu Road, Pudong New Area,
Shanghai 200135, China
E-mail: yangtieyi@163.com
First Author:
Shan Su
Affiliation: The Graduate School, Ningxia Medical University, Yinchuan, 750004, PR China.
Postal address: No.1160, Shengli Street, Xingqing District, Yinchuan, Ningxia Hui
Autonomous Region
E-mail: s18755142360@163.com
Background: Estimated pulse wave velocity (ePWV) serves as a noninvasive indicator of arteriosclerosis, reflecting the extent of vascular aging. Previous studies have demonstrated a negative correlation between arteriosclerosis and bone mineral density (BMD), with reduced BMD recognised as a primary marker of osteoporosis. Nevertheless, the relationship between ePWV and osteoporosis risk, and the mechanisms connecting arteriosclerosis to osteoporosis in postmenopausal women, remain unclear.
Methods: This study included 6,281 postmenopausal women aged ≥50 years who were initially free from osteoporosis, using data obtained from the Health and Retirement Study (HRS) conducted in the United States. The ePWV was calculated based on age and mean arterial pressure (MBP). Subgroup analyses were performed according to demographic and clinical characteristics. A Cox proportional hazards model was employed to evaluate the association between ePWV and osteoporosis risk in postmenopausal women. Results were presented as hazard ratios (HRs) with corresponding 95% confidence intervals (CIs).
Results: Elevated ePWV was associated with a heightened osteoporosis risk among postmenopausal women. After stratifying participants using X-tile cutoffs, compared with participants who had an ePWV of <9.04 m/s, the osteoporosis risk was significantly elevated by 25% (HR = 1.25, 95% CI: 1.08–1.43, P = 0.002) in the group with ePWV of 9.04–12.25 m/s, and by 39% (HR = 1.39, 95% CI: 1.17–1.66, P < 0.001) in the group with ePWV of ≥12.25 m/s.
Conclusion: The findings demonstrate that higher ePWV is associated with increased osteoporosis risk in postmenopausal women, highlighting ePWV’s potential utility for early risk stratification. These results require further validation in external cohort studies.
Keywords: HRS; ePWV; Arteriosclerosis; Osteoporosis; Postmenopausal women
Osteoporosis, a common systemic skeletal disorder characterised by increased bone fragility and fracture risk, is associated with elevated mortality. Evidence has demonstrated that patients with osteoporosis exhibit a significantly greater risk of cardiovascular disease (CVD) and mortality compared to the general population [1,2]. Estrogen deficiency-related postmenopausal bone loss is the primary cause of osteoporosis, which is prevalent among postmenopausal women, with prevalence steadily rising due to population ageing [3]. Therefore, proactively identifying risk factors for osteoporosis is crucial for its prevention and reducing associated disease burdens.
Both atherosclerosis and bone loss are age-related processes influenced by common underlying factors, including oxidative stress, inflammation, and hormonal imbalances [4]. In addition, osteoporosis and vascular sclerosis frequently co-occur, as vascular calcification and bone metabolism share common pathways involved in bone formation and mineralization [5]. Cross-sectional studies have indicated that the association between brachial-ankle pulse wave velocity (baPWV), a widely used marker of arteriosclerosis, and bone mineral density (BMD) and osteoporosis differs by gender, with elevated baPWV typically correlating negatively with BMD in women [6]. In postmenopausal women, baPWV levels were higher in individuals with osteoporosis compared to those with osteopenia or normal bone mass, indicating a negative correlation between baPWV and femoral BMD [7]. Estimated pulse wave velocity (ePWV), calculated from age and blood pressure, serves as a simple surrogate indicator of arteriosclerosis, with predictive capabilities comparable or even superior to baPWV and carotid-femoral pulse wave velocity (cfPWV) for assessing CVD and mortality risk [8,9]. Moreover, several cohort studies have shown that elevated ePWV is significantly associated with an increased risk of diabetes, dementia, and mortality [10–12]. However, longitudinal evidence regarding the relationship between ePWV and osteoporosis risk remains limited. Consequently, further validation through cohort studies is necessary to clarify the association between arteriosclerosis and osteoporosis risk in postmenopausal women.
This study aims to investigate the relationship between ePWV and osteoporosis risk in postmenopausal women, hypothesising that ePWV may serve as a comprehensive biomarker reflecting adverse systemic conditions associated with bone loss. Data from the Health and Retirement Study (HRS) were utilised to offer insights into early risk stratification and highlight vascular assessment as a potential tool for osteoporosis prevention.
Study Design
The analysis was conducted using data from the HRS in the United States, which employs a multistage area probability sampling design with oversampling of Black American adults and their families. This study used data from the 2012 and 2014 waves as the baseline, and participants were subsequently followed up every two years until 2022. Participants underwent interviews regarding demographics, environmental factors, social connections, and household composition, which were integrated with assessments of physical health, functional status, and disability indicators. This study included non-proxy postmenopausal women aged ≥50 years from the 2012 and 2014 waves who had complete data on key biomarkers, including age, systolic blood pressure, diastolic blood pressure, grip strength, PEF, and cystatin C, which were used for ePWV calculation or covariate assessment, resulting in an initial sample size of 9,088 participants. The follow-up period was calculated from the baseline interview to the first reported diagnosis of osteoporosis, which was defined as the event. Participants who did not develop osteoporosis were censored at their last follow-up visit, and those lost to follow-up were censored at the time of their last completed interview. Participants were excluded if they met any of the following criteria: (1) osteoporosis at baseline; (2) incomplete data for ePWV assessment at baseline; (3) premenopausal status; or (4) absence of osteoporosis assessment during follow-up. Finally, 6,281 participants were included in the analysis.
Data extraction and screening
Flow chart of the data screening process:
Figure 1 Flow chart of data screening
Variable Definitions and Handling
Outcome variables:
The HRS collects health information biennially through telephone or face-to-face interviews. The osteoporosis outcome was determined based on participants’ self-reports of whether they had “ever been diagnosed with osteoporosis by a physician”, corresponding to questionnaire items NC280, OC280, PC280, QC280, RC280, and SC280.
Primary exposure of interest:
Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured three consecutive times in a seated position, with intervals of 45 seconds between measurements. The average of the three readings was calculated and used in the analysis. ePWV (m/s) was computed using the following formula:
ePWV[9] = 9.587 − 0.402 × Age + 4.560 × 10-3 × Age2 − 2.621 × 10-5 × Age2 × MBP + 3.176 × 10-3 × Age × MBP − 1.832 × 10-2 × MBP
In this equation, age is expressed in consecutive years, and mean arterial pressure (MBP) is calculated as:
MBP = DBP + 0.4 × (SBP－DBP)
X-tile software was used in this study to exploratively identify optimal ePWV thresholds for categorising participants, using osteoporosis occurrence as the endpoint. Based on these thresholds, ePWV was categorised into three groups: <9.04 m/s, 9.04–12.25 m/s, and ≥12.25 m/s.
Covariates:
The established models controlled for factors associated with osteoporosis risk, including sociodemographic characteristics, health behaviours, BMI, health status, CVD, and medication use. Sociodemographic variables included race/ethnicity (Caucasian, African American, others), education level (high school or below, above high school), and marital status (married, unmarried). Health behaviour variables comprised physical activity (multiple times per week, once per week, 1–3 times per month, rarely or never), current smoking (yes/no), current drinking (yes/no), diabetes mellitus (yes/no), CVD (yes/no), hip fracture (yes/no), and hypertensive medication use (yes/no). In addition to these self-reported indicators, this study also included BMI (kg/m2), PEF (L/min), low muscle strength (kg), and laboratory-measured cystatin C levels (mg/L, as an indicator of overall kidney dysfunction). Based on previous studies, clinical significance, and univariate analysis results, the multivariate model included race, diabetes mellitus, CVD, BMI, PEF, low muscle strength, and cystatin C for adjustment.
Definitions of key variables:
(1) Diabetes status was determined using self-reported medical history and laboratory test results (HbA1c ≥6.5% or self-reported physician diagnosis).
(2) CVD was defined based on self-reported physician diagnoses from the HRS questionnaire, including myocardial infarction, coronary heart disease/ischemic heart disease, heart failure, stroke, and other related cardiovascular and cerebrovascular diseases.
(3) PEF (L/min) was defined as the maximum value from three measurements, taken with intervals of at least 30 seconds.
(4) Grip strength was measured alternately on both hands, with intervals of at least 15 seconds between three measurements per hand. The maximum value was used as the final result. If grip strength could not be measured on one hand, the maximum value from the other hand was recorded. Low muscle strength was defined as grip strength <16 kg according to FNIH criteria [13].
Statistical Analysis Methods
Descriptive statistics:
Quantitative data were assessed for normality using the Shapiro-Wilk test and for homogeneity of variances using Levene’s test. Normally distributed quantitative data were expressed as the mean ± standard deviation (Mean ± SD). The t-test was employed for between-group comparisons when variances were homogeneous, whereas the t’-test was used when variances were heterogeneous. Non-normally distributed data were presented as the median and interquartile range [M (Q1, Q3)] and compared between groups using the Wilcoxon rank-sum test. Categorical data were expressed as frequencies and percentages [n (%)] and compared using the chi-square test or Fisher’s exact test.
Univariate analysis:
Initially, a univariate Cox proportional hazards model was utilised to evaluate associations between each variable and the risk of osteoporosis. Based on the univariate analysis results, prior literature, and clinical significance, relevant variables were selected for inclusion in a multivariate model to examine the independent relationship between ePWV and osteoporosis risk. Additionally, the variance inflation factor (VIF) for independent variables in the multivariate model was calculated to assess multicollinearity.
Subgroup analyses:
Subgroup analyses were performed based on demographic and clinical characteristics to evaluate the discriminative capability of ePWV for osteoporosis risk across populations with differing features.
Evaluation index: Hazard ratio (HR) and 95% confidence interval (CI).
Confidence level and statistical software: Statistical significance was set at α = 0.05. All data extraction and analyses were conducted using R software (version 4.5.1).
Results
Ⅰ. Comparison of baseline characteristics
This study included 6,281 participants (Table 1), divided into two groups based on osteoporosis status: 5,009 participants (79.76%) in the non-osteoporosis group and 1,272 participants (20.24%) in the osteoporosis group. Comparisons of continuous variables between groups revealed that the mean follow-up duration (years) was significantly shorter in the osteoporosis group compared to the non-osteoporosis group (4.82 ± 2.54 vs. 7.30 ± 2.55, *t*’ = 31.037, P < 0.001). Mean age was significantly higher in the osteoporosis group than in the non-osteoporosis group (66.91 ± 9.60 vs. 65.36 ± 10.33, *t*’ = -5.039, P < 0.001). Additionally, MBP was significantly lower in the osteoporosis group than in the non-osteoporosis group (96.52 ± 13.77 vs. 98.62 ± 14.20, *t* = 4.741, P < 0.001).
The comparison of categorical variables between groups revealed a statistically significant difference in the distribution of ePWV categories (χ2 = 9.131, P = 0.010). Specifically, the proportion of participants with ePWV ≥12.25 m/s was slightly higher in the osteoporosis group compared with the non-osteoporosis group (25.08% vs. 22.78%), while the proportion with ePWV <9.04 m/s was slightly lower (23.51% vs. 27.55%). Significant differences between groups were observed in sociodemographic characteristics (χ2 = 62.168, P < 0.001). The osteoporosis group had a significantly higher proportion of Caucasian participants (77.59% vs. 66.50%), a significantly lower proportion of African American participants (15.41% vs. 25.19%), and a slightly lower proportion of participants from other racial groups (7.00% vs. 8.31%). In terms of education level, the proportion of participants with at least a high school diploma was significantly higher in the osteoporosis group compared to the non-osteoporosis group (52.52% vs. 48.53%, χ2 = 6.280, P = 0.012). Regarding chronic disease history, the prevalence of diabetes mellitus was significantly lower in the osteoporosis group than in the non-osteoporosis group (20.44% vs. 25.59%, χ2 = 14.269, P < 0.001). In terms of medication use, the proportion of patients taking antihypertensive drugs was significantly lower in the osteoporosis group compared with the non-osteoporosis group (51.89% vs. 55.12%, χ2 = 4.148, P = 0.042). Regarding body mass index (BMI) classification, the proportion of participants with a BMI <25 kg/m2 was significantly higher in the osteoporosis group (13.60% vs. 9.56%), whereas the proportion with a BMI ≥25 kg/m2 was significantly lower (86.40% vs. 90.44%; χ2 = 17.347, P < 0.001).
Table 1 Comparison of baseline characteristics
| Variables | Total (N=6281) |
Osteoporosis_2=No (N=5009) |
Osteoporosis_2=Yes (N=1272) |
Statistics | P |
| Time(years), Mean (±SD) | 6.80 (±2.74) | 7.30 (±2.55) | 4.82 (±2.54) | t’ = 31.037 | <0.001 |
| ePWV(m/s), Mean (±SD) | 10.64 (±2.22) | 10.62 (±2.24) | 10.70 (±2.17) | t = -1.140 | 0.254 |
| ePWV groups, n (%) | χ² = 9.131 | 0.010 | |||
| <9.04 | 1679 (26.73) | 1380 (27.55) | 299 (23.51) | ||
| 9.04-12.25 | 3142 (50.02) | 2488 (49.67) | 654 (51.42) | ||
| ≥12.25 | 1460 (23.24) | 1141 (22.78) | 319 (25.08) | ||
| Age(years), Mean (±SD) | 65.68 (±10.20) | 65.36 (±10.33) | 66.91 (±9.60) | t’ = -5.039 | <0.001 |
| Race, n (%) | χ² = 62.168 | <0.001 | |||
| White/Caucasian | 4318 (68.75) | 3331 (66.50) | 987 (77.59) | ||
| Black or African American | 1458 (23.21) | 1262 (25.19) | 196 (15.41) | ||
| Others | 505 (8.04) | 416 (8.31) | 89 (7.00) | ||
| Education, n (%) | χ² = 6.280 | 0.012 | |||
| High School and Below | 3182 (50.66) | 2578 (51.47) | 604 (47.48) | ||
| High School and Above | 3099 (49.34) | 2431 (48.53) | 668 (52.52) | ||
| Marital status, n (%) | χ² = 3.081 | 0.079 | |||
| Married | 3341 (53.19) | 2636 (52.63) | 705 (55.42) | ||
| Not married | 2940 (46.81) | 2373 (47.37) | 567 (44.58) | ||
| Vigorous physical, n (%) | χ² = 1.483 | 0.686 | |||
| More than once a week | 1415 (22.53) | 1137 (22.70) | 278 (21.86) | ||
| Once a week | 634 (10.09) | 500 (9.98) | 134 (10.53) | ||
| One to three times a month | 589 (9.38) | 461 (9.20) | 128 (10.06) | ||
| Hardly ever or never | 3643 (58.00) | 2911 (58.12) | 732 (57.55) | ||
| Moderate physical, n (%) | χ² = 4.415 | 0.220 | |||
| More than once a week | 3038 (48.37) | 2405 (48.01) | 633 (49.76) | ||
| Once a week | 1061 (16.89) | 871 (17.39) | 190 (14.94) | ||
| One to three times a month | 842 (13.41) | 670 (13.38) | 172 (13.52) | ||
| Hardly ever or never | 1340 (21.33) | 1063 (21.22) | 277 (21.78) | ||
| Smoking, n (%) | χ² = 3.411 | 0.065 | |||
| No | 3220 (51.27) | 2538 (50.67) | 682 (53.62) | ||
| Yes | 3061 (48.73) | 2471 (49.33) | 590 (46.38) | ||
| Drinking, n (%) | χ² = 0.765 | 0.382 | |||
| No | 3029 (48.22) | 2430 (48.51) | 599 (47.09) | ||
| Yes | 3252 (51.78) | 2579 (51.49) | 673 (52.91) | ||
| Diabetes, n (%) | χ² = 14.269 | <0.001 | |||
| No | 4739 (75.45) | 3727 (74.41) | 1012 (79.56) | ||
| Yes | 1542 (24.55) | 1282 (25.59) | 260 (20.44) | ||
| CVD, n (%) | χ² = 1.856 | 0.173 | |||
| No | 4920 (78.33) | 3942 (78.70) | 978 (76.89) | ||
| Yes | 1361 (21.67) | 1067 (21.30) | 294 (23.11) | ||
| Hip fractured, n (%) | χ² = 0.300 | 0.584 | |||
| No | 6250 (99.51) | 4986 (99.54) | 1264 (99.37) | ||
| Yes | 31 (0.49) | 23 (0.46) | 8 (0.63) | ||
| Anti–hypertension drugs, n (%) | χ² = 4.148 | 0.042 | |||
| No | 2860 (45.53) | 2248 (44.88) | 612 (48.11) | ||
| Yes | 3421 (54.47) | 2761 (55.12) | 660 (51.89) | ||
| BMI(kg/m2), n (%) | χ² = 17.347 | <0.001 | |||
| <25 | 652 (10.38) | 479 (9.56) | 173 (13.60) | ||
| ≥25 | 5629 (89.62) | 4530 (90.44) | 1099 (86.40) | ||
| PEF(L/min), Mean (±SD) | 5.37 (±1.51) | 5.38 (±1.52) | 5.35 (±1.49) | t = 0.767 | 0.443 |
| Low muscular strength, n (%) | χ² = 0.561 | 0.454 | |||
| No | 5734 (91.29) | 4580 (91.44) | 1154 (90.72) | ||
| Yes | 547 (8.71) | 429 (8.56) | 118 (9.28) | ||
| MBP(mmHg), Mean (±SD) | 98.20 (±14.14) | 98.62 (±14.20) | 96.52 (±13.77) | t = 4.741 | <0.001 |
| Cystatin–C(Mg/L), M (Q₁, Q₃) | 0.66 (0.53, 0.82) | 0.66 (0.53, 0.82) | 0.66 (0.53, 0.82) | W = 3150553.000 | 0.543 |
| Note: SD: Standard Deviation; M: Median; Q₁: 1st Quartile; Q₃: 3rd Quartile t: Student’s t test; t’: Satterthwaite t test; W: Wilcoxon rank sum test; χ²: Chi-square test; -: Fisher’s exact test F: Analysis of Variance; F’: Welch Anova test; H: Kruskal-Wallis H test |
|||||
- Covariate Screening Process
- Covariate screening using a univariate Cox proportional hazards model
A univariate Cox proportional hazards model was employed to assess associations between each variable and osteoporosis risk (Table 2). Variables with P < 0.05 were included in the subsequent multivariate analysis as covariates for adjustment. The final covariates included in the model were race, diabetes mellitus, CVD, BMI, PEF, low muscle strength, and cystatin C.
The results indicated that each one-unit increase in ePWV (as a continuous variable) significantly elevated osteoporosis risk by 7% (HR = 1.07, 95% CI: 1.04–1.09, P < 0.001). After categorising ePWV into tertiles, osteoporosis risk progressively increased in the 9.04–12.25 m/s group (HR = 1.26, 95% CI: 1.10–1.44, P = 0.001) and the ≥12.25 m/s group (HR = 1.63, 95% CI: 1.39–1.91, P < 0.001) compared with the <9.04 m/s group. Regarding race, osteoporosis risk was significantly lower among African American participants (HR = 0.54, 95% CI: 0.46–0.63, P < 0.001) and other racial groups (HR = 0.72, 95% CI: 0.58–0.90, P = 0.003) compared with Caucasians. Diabetes mellitus showed a significant negative association with osteoporosis risk (HR = 0.81, 95% CI: 0.71–0.93, P = 0.003), whereas a history of CVD was significantly associated with increased risk (HR = 1.23, 95% CI: 1.08–1.40, P = 0.002). Participants with BMI ≥25 kg/m2 had significantly reduced osteoporosis risk compared with those having BMI <25 kg/m2 (HR = 0.63, 95% CI: 0.54–0.74, P < 0.001). Higher PEF was significantly associated with decreased osteoporosis risk, with an 11% risk reduction per unit increase (HR = 0.89, 95% CI: 0.84–0.95, P < 0.001). Low muscle strength significantly increased osteoporosis risk (HR = 1.35, 95% CI: 1.12–1.63, P = 0.002). Finally, each unit increase in cystatin C significantly increased osteoporosis risk by 8% (HR = 1.08, 95% CI: 1.03–1.13, P = 0.002)
Table 2 Covariate screening process
| Variables | HR (95% CI) | P |
| ePWV , m/s | 1.07 (1.04-1.09) | <0.001 |
| ePWV groups | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.26 (1.10-1.44) | 0.001 |
| ≥12.25 | 1.63 (1.39-1.91) | <0.001 |
| Race | ||
| White/Caucasian | Ref | |
| Black or African American | 0.54 (0.46-0.63) | <0.001 |
| Others | 0.72 (0.58-0.90) | 0.003 |
| Education | ||
| High School and Below | Ref | |
| High School and Above | 1.06 (0.95-1.19) | 0.262 |
| Marital status | ||
| Married | Ref | |
| Not married | 0.96 (0.86-1.08) | 0.501 |
| Vigorous physical | ||
| More than once a week | Ref | |
| Once a week | 1.13 (0.92-1.39) | 0.231 |
| One to three times a month | 1.16 (0.94-1.43) | 0.162 |
| Hardly ever or never | 1.13 (0.99-1.30) | 0.078 |
| Moderate physical | ||
| More than once a week | Ref | |
| Once a week | 0.88 (0.75-1.03) | 0.114 |
| One to three times a month | 1.02 (0.86-1.20) | 0.852 |
| Hardly ever or never | 1.15 (1.00-1.32) | 0.054 |
| Smoking | ||
| No | Ref | |
| Yes | 0.94 (0.84-1.04) | 0.239 |
| Drinking | ||
| No | Ref | |
| Yes | 0.97 (0.87-1.08) | 0.557 |
| Diabetes | ||
| No | Ref | |
| Yes | 0.81 (0.71-0.93) | 0.003 |
| CVD | ||
| No | Ref | |
| Yes | 1.23 (1.08-1.40) | 0.002 |
| Hip fractured | ||
| No | Ref | |
| Yes | 1.86 (0.93-3.73) | 0.080 |
| Anti–hypertension drugs | ||
| No | Ref | |
| Yes | 0.94 (0.85-1.05) | 0.306 |
| BMI, kg/m2 | ||
| <25 | Ref | |
| ≥25 | 0.63 (0.54-0.74) | <0.001 |
| PEF, L/min | 0.89 (0.84-0.95) | <0.001 |
| Low muscular strength, kg | ||
| No | Ref | |
| Yes | 1.35 (1.12-1.63) | 0.002 |
| Cystatin–C, Mg/L | 1.08 (1.03-1.13) | 0.002 |
| Note: HR: Hazard ratio; CI: Confidence intervals; Ref: reference Single-factor model: unadjusted |
||
- Multivariate Cox proportional hazards model analysis
A multivariate Cox proportional hazards regression model was used to evaluate the association between ePWV (as continuous and categorical variables) and osteoporosis risk (Table 3). After adjusting for confounding factors, including race, diabetes mellitus, CVD, BMI, PEF, low muscle strength, and cystatin C, ePWV (continuous variable) remained significantly and positively associated with osteoporosis risk (HR = 1.03, 95% CI: 1.00–1.06, P = 0.039). When ePWV was categorised, the osteoporosis risk was significantly increased by 25% (HR = 1.25, 95% CI: 1.08–1.43, P = 0.002) and 39% (HR = 1.39, 95% CI: 1.17–1.66, P < 0.001) in the groups with ePWV levels of 9.04–12.25 m/s and ≥12.25 m/s, respectively, compared with the reference group (<9.04 m/s). Trend analysis indicated a significant positive correlation between rising ePWV levels and osteoporosis risk (P < 0.05).
Table 3 Results of multivariate Cox proportional hazards regression model
| Variables | HR (95% CI) | P |
| ePWV, m/s | 1.03 (1.00-1.06) | 0.039 |
| ePWV groups | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.25 (1.08-1.43) | 0.002 |
| ≥12.25 | 1.39 (1.17-1.66) | <0.001 |
| Note: HR: Hazard ratio; CI: Confidence intervals; Ref: reference Model adjustment: Race; Diabetes; CVD; BMI; PEF; Low muscular strength; Cystatin-C |
||
III. Exploration of the association between ePWV and osteoporosis risk in postmenopausal women
As shown in Table 4, elevated ePWV was identified as a significant risk factor for osteoporosis in postmenopausal women. Specifically, the osteoporosis risk was increased by 25% and 39% in the groups with ePWV levels of 9.04–12.25 m/s and ≥ 12.25 m/s, respectively, compared with the group with ePWV < 9.04 m/s. Additionally, the osteoporosis risk increased by 3% for each one-unit increase in ePWV.
Table 4 Results of correlation analysis
| Model 1 | Model 2 | |||
| Variables | HR (95% CI) | P | HR (95% CI) | P |
| ePWV, m/s | 1.07 (1.04-1.09) | <0.001 | 1.03 (1.00-1.06) | 0.039 |
| ePWV groups | ||||
| <9.04 | Ref | Ref | ||
| 9.04-12.25 | 1.26 (1.10-1.44) | 0.001 | 1.25 (1.08-1.43) | 0.002 |
| ≥12.25 | 1.63 (1.39-1.91) | <0.001 | 1.39 (1.17-1.66) | <0.001 |
Note: HR: Hazard ratio; CI: Confidence intervals; Ref: reference
Model 1: Unadjusted.
Model 2 Adjustments: Race; Diabetes; CVD; BMI; PEF; Low muscular strength; Cystatin C.
Figure 2 Relationship between ePWV and osteoporosis outcome
Figure 2(A) presents the relationship between ePWV and osteoporosis risk using restricted cubic spline (RCS) analysis. The findings illustrate a generally increasing trend in osteoporosis risk with elevated ePWV levels, without statistically significant nonlinear associations.
Figure 2(B) displays cumulative risk curves for ePWV categories defined by X-tile cutoff points. Participants with higher ePWV consistently demonstrated greater cumulative incidence of osteoporosis during follow-up, confirming significant differences among the three ePWV groups (P < 0.001). The most notable separation between curves was observed at approximately 5 years of follow-up, highlighting the prognostic value of ePWV in identifying individuals at high risk of osteoporosis.
- Subgroup Analysis
Predefined subgroup analyses were conducted based on age, race, antihypertensive medication use, diabetes mellitus history, CVD history, hip fracture history, BMI, and low muscle strength. Results are presented as forest plots (Figure 3). The positive association between ePWV and osteoporosis risk remained consistent across most subgroups. In particular, ePWV (as a continuous variable) was significantly associated with increased osteoporosis risk in subgroups defined by Caucasian race, antihypertensive medication use, absence of CVD, absence of hip fracture, BMI ≥25 kg/m2, and absence of low muscle strength.
Figure 3 Subgroup analysis of continuous ePWV
| Subgroups (Outcome/Total) |
HR (95% CI) | P |
| Age=<70(N=780/4074) | ||
| ePWV | 1.08 (0.96-1.22) | 0.197 |
| Age=≥70(N=492/2207) | ||
| ePWV | 0.94 (0.82-1.06) | 0.314 |
| Race=White/Caucasian(N=987/4318) | ||
| ePWV | 1.07 (1.00-1.15) | 0.043 |
| Race=Black or African American(N=196/1458) | ||
| ePWV | 1.08 (0.92-1.28) | 0.338 |
| Race=Others(N=89/505) | ||
| ePWV | 0.92 (0.69-1.21) | 0.535 |
| Anti–hypertension drugs=No(N=612/2860) | ||
| ePWV | 1.06 (0.97-1.16) | 0.219 |
| Anti–hypertension drugs=Yes(N=660/3421) | ||
| ePWV | 1.10 (1.01-1.21) | 0.031 |
| Diabetes=No(N=1012/4739) | ||
| ePWV | 1.05 (0.98-1.12) | 0.192 |
| Diabetes=Yes(N=260/1542) | ||
| ePWV | 1.15 (1.00-1.33) | 0.057 |
| CVD=No(N=978/4920) | ||
| ePWV | 1.08 (1.01-1.17) | 0.028 |
| CVD=Yes(N=294/1361) | ||
| ePWV | 1.02 (0.90-1.15) | 0.769 |
| Hip fractured=No(N=1264/6250) | ||
| ePWV | 1.07 (1.00-1.14) | 0.040 |
| BMI=<25(N=173/652) | ||
| ePWV | 1.06 (0.92-1.24) | 0.424 |
| BMI=≥25(N=1099/5629) | ||
| ePWV | 1.08 (1.01-1.15) | 0.035 |
| Low muscular strength=No(N=1154/5734) | ||
| ePWV | 1.08 (1.01-1.15) | 0.031 |
| Low muscular strength=Yes(N=118/547) | ||
| ePWV | 0.99 (0.83-1.19) | 0.952 |
| Note: HR: Hazard ratio; CI: Confidence intervals; Ref: reference Model adjustment: Race; Diabetes; CVD; BMI; PEF; Low muscular strength; Cystatin-C |
||
As illustrated in Figure 4, elevated ePWV was significantly associated with increased osteoporosis risk among subgroups defined by age (<70 and ≥70 years), Caucasian race, antihypertensive medication use (both yes and no), diabetes mellitus history (both yes and no), absence of CVD, absence of hip fracture, BMI ≥25 kg/m2, and absence of low muscle strength. Higher ePWV levels consistently corresponded to a greater risk of osteoporosis.
Figure 4 Subgroup analysis of categorical ePWV
| Subgroups (Outcome/Total) |
HR (95% CI) | P |
| Age<70years(N=780/4074) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.20 (1.04-1.39) | 0.015 |
| ≥12.25 | 1.64 (1.09-2.46) | 0.018 |
| Age≥70years(N=492/2207) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 357805.19 (0.00-Inf) | 0.990 |
| ≥12.25 | 366894.75 (0.00-Inf) | 0.990 |
| Race=White/Caucasian(N=987/4318) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.30 (1.10-1.53) | 0.002 |
| ≥12.25 | 1.45 (1.19-1.77) | <0.001 |
| Race=Black or African American(N=196/1458) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.19 (0.84-1.67) | 0.334 |
| ≥12.25 | 1.37 (0.88-2.15) | 0.167 |
| Race=Others(N=89/505) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.11 (0.72-1.73) | 0.630 |
| ≥12.25 | 0.77 (0.33-1.81) | 0.556 |
| Anti–hypertension drugs =No(N=612/2860) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.29 (1.08-1.55) | 0.005 |
| ≥12.25 | 1.32 (1.02-1.72) | 0.036 |
| Anti–hypertension drugs =Yes(N=660/3421) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.25 (0.99-1.58) | 0.065 |
| ≥12.25 | 1.52 (1.17-1.98) | 0.002 |
| Diabetes=No(N=1012/4739) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.25 (1.07-1.46) | 0.004 |
| ≥12.25 | 1.34 (1.11-1.63) | 0.003 |
| Diabetes=Yes(N=260/1542) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.24 (0.88-1.74) | 0.226 |
| ≥12.25 | 1.56 (1.04-2.33) | 0.032 |
| CVD=No(N=978/4920) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.26 (1.08-1.47) | 0.003 |
| ≥12.25 | 1.45 (1.19-1.76) | <0.001 |
| CVD=Yes(N=294/1361) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.19 (0.85-1.65) | 0.310 |
| ≥12.25 | 1.22 (0.84-1.78) | 0.288 |
| Hip fractured=No(N=1264/6250) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.24 (1.08-1.43) | 0.002 |
| ≥12.25 | 1.41 (1.18-1.68) | <0.001 |
| BMI<25kg(N=173/652) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.37 (0.94-1.99) | 0.098 |
| ≥12.25 | 1.36 (0.86-2.13) | 0.185 |
| BMI≥25kg(N=1099/5629) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.23 (1.06-1.44) | 0.006 |
| ≥12.25 | 1.41 (1.17-1.71) | <0.001 |
| Low muscular strength=No(N=1154/5734) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.25 (1.09-1.45) | 0.002 |
| ≥12.25 | 1.41 (1.18-1.70) | <0.001 |
| Low muscular strength=Yes(N=118/547) | ||
| ePWV_3 | ||
| <9.04 | Ref | |
| 9.04-12.25 | 1.03 (0.56-1.91) | 0.923 |
| ≥12.25 | 1.10 (0.59-2.04) | 0.774 |
| Note: HR: Hazard ratio; CI: Confidence intervals; Ref: reference Model adjustment: Race; Diabetes; CVD; BMI; PEF; Low muscular strength; Cystatin-C |
||
Discussion
This study identified a significant association between ePWV, a measure of arteriosclerosis, and an elevated risk of self-reported osteoporosis in a cohort of postmenopausal women. Each one-unit increase in ePWV corresponded to a 3% rise in osteoporosis risk, with no significant nonlinear associations observed. These findings align with previous studies that have demonstrated a negative correlation between arteriosclerosis and bone density [14,15]. Therefore, clinical assessments for cardiovascular risk in postmenopausal women with arteriosclerosis should also prioritise monitoring bone health.
Both osteoporosis and arteriosclerosis deteriorate with age and share common underlying pathophysiological mechanisms [16,17]. In osteoporosis, shared pathways involving inflammation, oxidative stress, and mineralisation regulators between bone metabolism and vascular calcification contribute to increased arterial stiffness [18–20]. Chronic inflammation and oxidative stress intensify progressively with age. Additionally, systemic pathological conditions, such as hormonal and metabolic disorders, might connect vascular structural changes with alterations in bone metabolism, suggesting the existence of a common “bone-vascular axis.”
Previous studies indicate that elevated oxidative stress represents another crucial mechanism underpinning damage to the “bone-vascular axis,” creating a vicious cycle through interactions with chronic inflammation in a causal feedback loop [21,22]. Oxidative stress describes a physiological state wherein the body’s production of pro-oxidant substances, primarily reactive oxygen species (ROS), exceeds the scavenging capacity of antioxidant defence systems, resulting in damage to cellular and molecular structures [23]. The pathological basis underlying elevated ePWV in this study involves structural vascular wall alterations caused by prolonged oxidative stress exposure [24]. Therefore, ePWV can be regarded as a functional reflection of cumulative oxidative vascular damage.
Model 2 in this study adjusted for diabetes history and metabolic and renal indicators, such as cystatin C. Even after these adjustments, the association between ePWV and osteoporosis risk persisted, suggesting a potentially independent relationship. The significant association after controlling for these variables implies that oxidative damage may exert its influence independently or through additional pathways. Patients with type 2 diabetes mellitus (T2DM) often exhibit poorer vascular health and earlier onset of vascular lesions, negatively impacting bone metabolism [25,26]. Abnormal interactions between advanced glycation end products (AGEs) and essential cytoplasmic factors can reduce vascular elasticity, enhance vascular stiffness, and exacerbate hypoxia [27]. Hypoxia may further stimulate pro-inflammatory factor and ROS production, consequently affecting bone metabolism [28]. In vascular smooth muscle cells, oxidised low-density lipoprotein (LDL) can induce cystatin C secretion, with moderate cystatin C levels inhibiting elastin degradation. However, persistently elevated cystatin C levels due to sustained oxidative stress can promote structural remodelling of the vascular adventitia, acting as a biological amplifier of oxidative damage [29–31].
For orthopaedic surgeons, patients with elevated ePWV levels may have bones chronically exposed to a microenvironment characterised by high oxidative stress. Even if bone density has not decreased below diagnostic thresholds, bone quality, including collagen integrity and cellular activity, may already be compromised, increasing fracture risk. This indicates that lifestyle management could benefit postmenopausal middle-aged and older women, although further validation through randomised trials is required.
Hormonal and metabolic signalling dysregulation is the primary initiator and amplifier of “bone-vascular axis” dysfunction in postmenopausal women [32–34]. This dysfunction does not arise from changes in a single hormone but rather from a network centred on estrogen deficiency, which triggers cascading imbalances across multiple downstream signalling pathways. In the present study, the independent association observed between ePWV and osteoporosis risk, even after adjusting for traditional risk factors, suggests a deeper shared pathophysiological foundation. We propose that systemic hormonal and metabolic dysregulation centred on estrogen deficiency constitutes this underlying mechanism.
The sharp decline in estrogen levels following menopause directly affects bone metabolism and impairs vascular endothelial function [35,36]. In addition, this altered microenvironment frequently involves insulin resistance, vitamin D deficiency or resistance, and abnormal parathyroid hormone (PTH) levels, collectively forming a metabolic background detrimental to both bone and vascular health [37–39]. Variables such as diabetes history, BMI, and cystatin C were adjusted in this study to control for confounding effects related to metabolic disturbances.
Therefore, elevated ePWV may represent cumulative damage from this metabolic network on the vascular wall. Moreover, bone tissue, as another highly metabolically active organ, may simultaneously experience exposure to this adverse environment. Clinically, postmenopausal women with elevated ePWV and normal bone density should be informed of their high metabolic turnover state and increased fracture risk, thus requiring comprehensive lifestyle and metabolic management strategies. Longitudinal studies are necessary to further clarify causal relationships and assess the efficacy of targeted interventions such as lifestyle modifications or pharmacological treatments for individuals exhibiting advanced vascular ageing and osteoporosis risk.
This study has certain limitations. First, it is an observational study, and although we made efforts to control for potential confounding factors, the association between estimated pulse wave velocity (ePWV) and the occurrence of osteoporosis in postmenopausal women cannot be directly interpreted as causal. Second, the outcome indicator for osteoporosis was based on patients’ self-reported physician diagnoses rather than standardized bone mineral density (BMD) measurements, which may introduce outcome classification bias and potentially attenuate the observed associations. Despite these limitations, this study extends prior knowledge regarding the importance of arterial stiffness.
Conclusion
In conclusion, this study identifies elevated ePWV as a significant risk factor for osteoporosis among postmenopausal women, with higher ePWV levels corresponding to increased osteoporosis risk. These findings preliminarily indicate the clinical potential of ePWV for early stratification of bone loss risk in postmenopausal women.
Acknowledgement
Tieyi Yang, and Dece Kong contributed to the conception and design of the work. Shan Su and Yiding Zhao made contributions to data collection, analysis, and interpretation. Shan Su、Yiding Zhao, and Dece Kong contributed to manuscript writing. Tieyi Yang, and Dece Kong contributed to manuscript polishing and material supplementation during the revision process. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.
Funding Sources
None.
Conflict of Interest
The author has no conflict of interest that requires disclosure.
Data availability
More information on the study design, questionnaires, and other details about the HRS can be found on the HRS website portal (http://hrsonline.isr.umich.edu/).
References
[1] Cai S, Fan J, Zhu L, et al. Bone mineral density and osteoporosis in relation to all-cause and cause-specific mortality in NHANES: A population-based cohort study[J]. Bone, 2020, 141:115597. DOI:10.1016/j.bone.2020.115597.
[2] Rodríguez-Gómez I, Gray S R, Ho F K, et al. Osteoporosis and Its Association With Cardiovascular Disease, Respiratory Disease, and Cancer: Findings From the UK Biobank Prospective Cohort Study[J]. Mayo Clinic Proceedings, 2022, 97(1): 110-121. DOI:10.1016/j.mayocp.2021.07.019.
[3] Michael R, JoAnn V, Felicia A, et al. Management of osteoporosis in postmenopausal women: the 2021 position statement of The North American Menopause Society[J]. Menopause, 2021, 28(9)973-997.
DOI:10.1097/GME.0000000000001831.
[4] Tap L, Kirkham F A, Mattace-Raso F, et al. Unraveling the Links Underlying Arterial Stiffness, Bone Demineralization, and Muscle Loss[J]. Hypertension, 2020, 76(3): 629-639. DOI:10.1161/HYPERTENSIONAHA.120.15184.
[5] Sage A P, Tintut Y, Demer L L. Regulatory Mechanisms in Atherosclerotic Calcification[J]. Nature reviews. Cardiology, 2010, 7(9): 528-536. DOI:10.1038/nrcardio.2010.115.
[6] Jannasz I, Brzeziński J, Mańczak M, et al. Is the association between pulse wave velocity and bone mineral density the same for men and women? – A systematic review and meta-analysis[J]. Archives of Gerontology and Geriatrics, 2024, 119: 105309. DOI:10.1016/j.archger.2023.105309.
[7] Seo S K, Cho S, Kim H Y, et al. Bone mineral density, arterial stiffness, and coronary atherosclerosis in healthy postmenopausal women[J]. Menopause, 2009, 16(5): 937-943. DOI:10.1097/gme.0b013e3181a15552.
[8] Hsu P, Lee W, Tsai W, et al. Comparison between estimated and brachial‐ankle pulse wave velocity for cardiovascular and overall mortality prediction[J]. The Journal of Clinical Hypertension, 2020, 23(1): 106-113. DOI:10.1111/jch.14124.
[9] Greve S V, Blicher M K, Kruger R, et al. Estimated carotid–femoral pulse wave velocity has similar predictive value as measured carotid–femoral pulse wave velocity[J]. Journal of Hypertension, 2016, 34(7): 1279-1289. DOI:10.1097/HJH.0000000000000935.
[10] Li D, Cao F, Cheng W, et al. Predictive value of estimated pulse wave velocity for cardiovascular and all-cause mortality in individuals with obesity[J]. Diabetology & Metabolic Syndrome, 2023, 15(1): 40. DOI:10.1186/s13098-023-01011-2.
[11] Heffernan K S, Wilmoth J M, London A S. Estimated Pulse Wave Velocity Is Associated With a Higher Risk of Dementia in the Health and Retirement Study[J]. American Journal of Hypertension, 2024, 37(11): 909-915. DOI:10.1093/ajh/hpae096.
[12] Zhang C, Chen L, Liu R. Elevated estimated pulse wave velocity and the risk of type 2 diabetes in non-obese young adults: a longitudinal cohort study[J]. BMC Endocrine Disorders, 2025, 25(1): 139. DOI:10.1186/s12902-025-01967-4.
[13] Studenski S A, Peters K W, Alley D E, et al. The FNIH Sarcopenia Project: Rationale, Study Description, Conference Recommendations, and Final Estimates[J]. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 2014, 69(5): 547-558. DOI:10.1093/gerona/glu010.
[14] Campos-Obando N, Kavousi M, Roeters van Lennep J, et al. Bone health and coronary artery calcification: The Rotterdam Study[J]. Atherosclerosis, 2015, 241(1): 278-283. DOI:10.1016/j.atherosclerosis.2015.02.013.
[15] Giallauria F, Ling S M, Schreiber C, et al. Arterial Stiffness and Bone Demineralization: The Baltimore Longitudinal Study of Aging[J]. American Journal of Hypertension, 2011, 24(9): 970-975. DOI:10.1038/ajh.2011.80.
[16] Szekanecz Z, Raterman H G, Pethő Z, et al. Common mechanisms and holistic care in atherosclerosis and osteoporosis[J]. Arthritis Research & Therapy, 2019, 21: 15. DOI:10.1186/s13075-018-1805-7.
[17] Avramovski P, Avramovska M, Sikole A. Bone Strength and Arterial Stiffness Impact on Cardiovascular Mortality in a General Population[J]. Journal of Osteoporosis, 2016, 2016: 7030272. DOI:10.1155/2016/7030272.
[18] Ranatunga S, Kulkarni B, Kinra S, et al. Sex-specific associations between markers of arterial stiffness and bone mineral density in Indian men and women[J]. Bone, 2023, 169: 116686. DOI:10.1016/j.bone.2023.116686.
[19] Pusztai A, Hamar A, Czókolyová M, et al. Associations of vascular and bone status in arthritis patients[J]. Scientific Reports, 2021, 11: 19504. DOI:10.1038/s41598-021-99071-9.
[20] Zhang H, Sun M, Hu Z, et al. Association of Arterial Stiffness with Functional Disability and Mobility Limitation: The Rugao Longitudinal Ageing Study[J]. The Journal of Nutrition, Health & Aging, 2022, 26(6): 598-605. DOI:10.1007/s12603-022-1802-z.
[21] Wang Y, Chen X, Xu J, et al. Association Between Arterial Stiffness and Bone Microarchitectural Deterioration in Type 2 Diabetes: A Cross‐Sectional Study[J]. Journal of Diabetes, 2025, 17(12): e70181. DOI:10.1111/1753-0407.70181.
[22] Ji J, Wu S, Bao X, et al. Mediating oxidative stress through the Palbociclib/miR-141-3p/STAT4 axis in osteoporosis: a bioinformatics and experimental validation study[J]. Scientific Reports, 2023, 13: 19560. DOI:10.1038/s41598-023-46813-6.
[23] Reddy V P. Oxidative Stress in Health and Disease[J]. Biomedicines, 2023, 11(11): 2925. DOI:10.3390/biomedicines11112925.
[24] González P, Lozano P, Ros G, et al. Hyperglycemia and Oxidative Stress: An Integral, Updated and Critical Overview of Their Metabolic Interconnections[J]. International Journal of Molecular Sciences, 2023, 24(11): 9352. DOI:10.3390/ijms24119352.
[25] Fajardo R J. Is Diabetic Skeletal Fragility Associated with Microvascular Complications in Bone?[J]. Current Osteoporosis Reports, 2017, 15(1): 1-8. DOI:10.1007/s11914-017-0341-8.
[26] Bora, S A. Carbonyl stress in diabetics with acute coronary syndrome[J]. Clinica Chimica Acta, 2021, 520: 78-86. DOI:10.1016/j.cca.2021.06.002.
[27] Xu J, Chen L, Yu H, et al. Involvement of Advanced Glycation End Products in the Pathogenesis of Diabetic Retinopathy[J]. Cellular Physiology and Biochemistry, 2018, 48(2): 705-717. DOI:10.1159/000491897.
[28] Malkov M I, Lee C T, Taylor C T. Regulation of the Hypoxia-Inducible Factor (HIF) by Pro-Inflammatory Cytokines[J]. Cells, 2021, 10(9): 2340. DOI:10.3390/cells10092340.
[29] Gao T, Liu F, Ban B, et al. Association between the ratio of serum creatinine to cystatin C and bone mineral density in Chinese older adults patients with type 2 diabetes mellitus[J]. Frontiers in Nutrition, 2022, 9: 1035853. DOI:10.3389/fnut.2022.1035853.
[30] Shao L, Chen S, Wu Y, et al. Association of serum urea-to-creatinine and creatinine-to-cystatin C ratios with bone mineral density and osteoporosis: a cross-sectional study[J]. Scientific Reports, 2025, 15: 27774. DOI:10.1038/s41598-025-13037-9.
[31] Quetglas-Llabrés M M, Díaz-López A, Bouzas C, et al. Association Between Oxidative–Inflammation Biomarkers and Incident Chronic Kidney Disease in People with High Cardiovascular Risk: A Nested Case–Control Study[J]. Antioxidants, 2025, 14(8): 975. DOI:10.3390/antiox14080975.
[32] Houben E, Neradova A, Schurgers L, et al. The influence of phosphate, calcium and magnesium on matrix Gla-protein and vascular calcification: a systematic review[J]. G Ital Nefrol, 2016, 33(6) : gin/33.6.5.
[33] Martini N, Streckwall L, McCarthy A D. Osteoporosis and vascular calcifications[J]. Endocrine Connections, 2023, 12(11): e230305. DOI:10.1530/EC-23-0305.
[34] Gebre A K, Lewis J R, Leow K, et al. Abdominal Aortic Calcification, Bone Mineral Density, and Fractures: A Systematic Review and Meta-analysis of Observational Studies[J]. The Journals of Gerontology: Series A, 2023, 78(7): 1147-1154. DOI:10.1093/gerona/glac171.
[35] Ji W, Pan B, Chen X, et al. Mendelian randomization studies of risk and protective factors for osteoporosis: a systematic review and meta-analysis[J]. Frontiers in Endocrinology, 2025, 15: 1486188. DOI:10.3389/fendo.2024.1486188.
[36] Iantomasi T, Romagnoli C, Palmini G, et al. Oxidative Stress and Inflammation in Osteoporosis: Molecular Mechanisms Involved and the Relationship with microRNAs[J]. International Journal of Molecular Sciences, 2023, 24(4): 3772. DOI:10.3390/ijms24043772.
[37] Molinuevo M S, Fernández J M, Cortizo A M, et al. Advanced glycation end products and strontium ranelate promote osteogenic differentiation of vascular smooth muscle cells in vitro: Preventive role of vitamin D[J]. Molecular and Cellular Endocrinology, 2017, 450: 94-104. DOI:10.1016/j.mce.2017.04.022.
[38] Gutierrez Y, Shaju R A, Sepulveda A, et al. Exploring the Interplay Between Hypertension and Osteoporosis: A Narrative Review[J]. Cureus, 2025, 17(11): e97459. DOI:10.7759/cureus.97459.
[39] Kim D, Piao J, Park J S, et al. Substance P-Mediated Vascular Protection Ameliorates Bone Loss[J]. Oxidative Medicine and Cellular Longevity, 2023, 2023: 9903336. DOI:10.1155/2023/9903336.