Abstract
Background
Postmenopausal osteoporosis (PMOP) has high morbidity and devastating complications. However, few studies explored the modifying role of different types of physical activity (PA) in the associations between the number of pregnancies as well as livebirths and the odds of PMOP. This study aimed to investigate the associations of the number of pregnancies and livebirths with PMOP odds.
Methods
This study used the baseline of the China Multi-Ethnic Cohort study, and a total of 27,459 postmenopausal women were recruited. Bone mineral density and quantitative ultrasound index were measured by quantitative ultrasound bone densitometer. Number of pregnancies and number of livebirths were collected by questionnaires. PA was classified into occupational, transportation, housework, and leisure PA. Logistic regression and multiple linear regression were used to analyze the associations between the number of pregnancies and livebirths and PMOP odds. The modifying effects of PA and its domain-specific types in the associations were analyzed by interaction analysis.
Results
A total of 1,539 (5.6%) participants had PMOP. Each additional pregnancy and each additional livebirth were associated with a 5% (OR [95% CI]: 1.05 [1.02, 1.09]) and 21% (1.21 [1.15, 1.28]) higher odds of PMOP, respectively. High levels of housework PA attenuated the adverse effect of livebirths on the PMOP odds (P for interaction < 0.05, q < 0.05). Associations between the number of pregnancies and livebirths and the odds of PMOP were observed among women of older age, with earlier age at menopause, later age at menarche, and longer total breastfeeding duration.
Conclusions
An increased number of pregnancies and livebirths was positively associated with the odds of PMOP. However, high levels of housework PA attenuated the adverse effect of livebirths on the PMOP odds, suggesting that an active lifestyle is associated with more favorable bone health status.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-026-27228-0.
Keywords: Postmenopausal osteoporosis, Postmenopausal woman, Pregnancy, Livebirth, Physical activity
Introduction
Osteoporosis is a systemic skeletal condition characterized by low bone mass and microarchitectural deterioration of bone tissue that increases bone fragility and fracture risk [1]. It is estimated that approximately over 200 million people suffered from osteoporosis worldwide from 2008 to 2018, and the incidence of fractures and osteoporosis among women in China was estimated to be more than three times higher than that among men [2–4]. Approximately 50% of postmenopausal women worldwide were affected by osteoporosis, and those with postmenopausal osteoporosis (PMOP) experience high morbidity and severe complications [5, 6]. Therefore, it is crucial to investigate the factors associated with PMOP.
Previous studies have suggested that the number of pregnancies and livebirths are associated with the odds of PMOP. Owing to the calcium demands of fetal bone development, pregnancy may strain maternal calcium reserves [7]. During pregnancy, 2% to 3% of the mother’s calcium, equivalent to approximately 25 to 30 g of calcium, is transferred to the fetus [8]. Breastfeeding further increases bone loss due to parathyroid hormone-related protein secretion and estrogen deficiency [9, 10]. However, previous studies have reported inconsistent results. For example, a study of 509 postmenopausal women in Turkey reported that spinal bone density decreased by 0.23 g/cm3 with each additional pregnancy [11]. Two other studies by Cure and Grainge suggested that the number of livebirths was associated with higher bone mineral density (BMD) in postmenopausal women [12, 13]. Therefore, the associations of the number of pregnancies and livebirths with PMOP odds remain inconclusive. Their associations are needed to validate in a large-scale population.
Regular physical activity (PA) could help prevent osteoporosis [14], while intensity of PA and the domain-specific PA, such as occupational, transportation, housework, and leisure, play different roles in bone health. A cohort study conducted in the northeast China found that the total PA, occupational PA, and housework PA were associated with lower odds of osteopenia in adults, while transportation and leisure PA were not [15]. A cross-sectional study found that PA ranging from moderate to vigorous was associated with a higher total spine BMD in postmenopausal women [16]. Therefore, the intensity of PA and domain-specific PA may attenuate the associations between these reproductive factors and PMOP. However, few studies have focused on the modifying effects of different types of PA on the associations of the number of pregnancies and livebirths with the odds of PMOP.
To address the research gaps, this study used the China Multi-Ethnic Cohort (CMEC) [17], which is one of the largest prospective cohorts in China. This cross-sectional study aimed to investigate the associations of the number of pregnancies and livebirths with the odds of PMOP, and the modifying effects of domain-specific PA in the associations. However, the findings would provide evidence for health stakeholders to develop public health intervention strategies for preventing PMOP.
Materials and methods
Study design and population
This study was based on the baseline of the CMEC, which included 98,609 adults aged 30–79 years from five southwest provinces of China between May 2018 and September 2019. Participants in CMEC were enrolled by a multistage, stratified cluster sampling method [17]. In this study, we excluded participants who (1) were men (n = 39,338); (2) lacked information on pregnancies and livebirths on the day of the survey (n = 181); (3) were not in menopause for at least 12 months (n = 30,529); (4) lacked information on bone health assessment status (n = 1,068); and (5) had self-reported osteoporosis (n = 34), since medication for osteoporosis taken by them may make their measurements not capable of reflecting the natural levels. Finally, a total of 27,459 participants from CEMC were included in this study (Figure S1). Each participant was interviewed in person with electronic questionnaire, and received physical examination data. The detailed content of the questionnaire employed in this study has been published previously [17]. The relevant content of questionnaire was in Table S12. All the participants signed an informed consent form before data collection. Ethical approval for this study was obtained from the Sichuan University Medical Ethical Review Board (K2016038).
Measurement of bone mineral and postmenopausal osteoporosis
Ultrasonic BMD measurements were conducted at the calcaneus of participants using a quantitative ultrasound bone densitometer (OSTEOKJ3000, KeJin Inc., Nanjing, China). Quantitative ultrasound bone densitometer is recognized as a validated methodology under the World Health Organization’s definition of osteoporosis, is a widely used tool for bone assessment in large population cohort [18]. Broadband ultrasound attenuation (BUA) and speed of sound (SOS) were used to estimate and calculate the ultrasonic BMD and quantitative ultrasound index (QUI). The ultrasonic BMD = 2.5926×(BUA + SOS)-3.687, unit mg/cm3; QUI = 0.41×(BUA + SOS)-571, no unit. The BMD and QUI reflect both bone quality and bone strength [19]. In this study, PMOP was diagnosed according to the 2018 Chinese guideline for the diagnosis and treatment of senile osteoporosis, which defines osteoporosis as an ultrasonic BMD value ≤ 2.5 standard deviations (SD) below the mean value for healthy individuals of the same sex [20]. According to previous studies, based on the T-score of calcaneus BMD measured by bone densitometer, participants were divided into normal bone mass (T >-2.5) and osteoporosis (T ≤-2.5) groups [21, 22].
Number of pregnancies and livebirths
The number of pregnancies referred to the total number of pregnancies a woman had experienced. The number of livebirths was defined as the number of newborns who exhibit at least one of the four vital signs of heartbeat, breathing, umbilical cord pulsation, and voluntary muscle contraction after delivery at or beyond 28 weeks of gestation. Based on previous study and the Chinese fertility situation, the number of pregnancies and livebirths were divided into 0, 1–2, 3–4, and ≥ 5 times, with the 1–2 times serving as the reference group [23, 24].
Measurement of physical activity
PA was assessed by the cumulative time participants spent engaging in their occupational, transportation, housework, and leisure PA in the past year. The intensity of PA was evaluated using metabolic equivalent (MET) [25]. METs measured the energy expenditure per hour and per kilogram of body weight for a specific domain of PA compared to that while sitting quietly, and one MET was defined as the energy expenditure while sitting for one hour [26]. The MET values for different types of activities were assigned based on the 2011 Compendium of PA, shown in Table S1 [25]. An individual’s daily PA level was calculated by multiplying the MET value of the activity by the number of hours spent on that activity per day. Total PA was calculated as the sum of daily MET values across occupational, transportation, housework, and leisure activities. Total PA, and domain-specific PA were categorized into low and high levels by the median value in this population (18.48 MET-h/day for total PA, 3.6 MET-h/day for occupational PA, 0 MET-h/day for transportation PA, 5.6 MET-h/day for housework PA, and 0 MET-h/day for leisure PA). Total PA was categorized into population‑specific quartiles among individuals with PMOP.
Measurement of covariates
Based on previous studies and a directed acyclic graph (DAG; Figure S2), we selected a set of variables for adjustment, comprising those that affect both the exposure and the outcome, as well as those that only influence the outcome [27, 28]. Sociodemographic characteristics included age, residential location (urban or rural), ethnicity (Han, minority), annual household income (< 12,000, 12,000–19,999, 20,000–59,999, ≥ 60,000 RMB), working environment (outdoor, indoor, others), education level (illiteracy, primary school, junior high school, high school, junior college or higher). Health-related behaviors included smoking status (no, yes, quit), diet (Alternate mediterranean diet [AMED]) [29], vitamin D or calcium supplements, and calcium intake. Reproductive factors included age at menopause, age at menarche, history of contraceptive use, history of hysterectomy or oophorectomy, and total breastfeeding duration.
Statistical analysis
Categorical variables were described as frequencies, and continuous variables were described as mean ± SD. Univariate analysis of covariates was performed by Analysis of Variance (ANOVA) or chi-square test. Logistic regression was used to describe the associations of the number and types of pregnancies and livebirths with PMOP, after adjusting for covariates. To test for linear trends, ordinal variables were included in the binary logistic regression model as continuous variables.
The modifying effects of total and domain-specific PA on the associations of the number of pregnancies and livebirths with the odds of PMOP were evaluated using an interaction term, with the significance tested via the likelihood ratio test to compare model fit with and without interaction terms. False discovery rate (FDR) correction was applied, and q < 0.05 was considered statistically significant for the interaction terms. The absolute risk differences (ARDs) were derived from the model-predicted probabilities, representing the change in PMOP prevalence associated with each additional pregnancy and livebirth, stratified by PA level. A series of stratified analyses were performed to examine the potential modifying effects of demographic and lifestyle factors, including age (≤ 60, > 60 years old), age at menopause (≤ 45, > 45 years old), age at menarche (< 15, ≥ 15 years old), and total breastfeeding duration (< 24, ≥ 24 months). Effect modification was tested using an interaction term of a subgroup variable and an exposure variable.
To evaluate the robustness of the results, five sensitivity analyses were conducted. First, participants with a self-reported history of fractures were excluded. Second, those who had never been pregnant were excluded. Third, we adjusted for environmental factors, including ultraviolet radiation B (UVB), temperature, and humidity. Fourth, BUA and SOS were used as continuous outcomes. Fifth, the classification criteria for adults’ weekly PA qualification recommended by the WHO were used for stratification.
All statistical analyses were performed by using R 4.0.2 software (R Foundation for Statistical Computing). Two-sided test with P value less than 0.05 were considered statistically significant.
Results
Characteristics of the included participants
Of 27,459 postmenopausal women 1,539 (5.60%) had PMOP. On average, women had 3.4 ± 1.7 pregnancies and 2.3 ± 1.3 livebirths. The mean total PA was 23.6 ± 17.9 MET-h/day. Women with PMOP were generally Han ethnics (61.7%), and indoor workers (44.4%), and were more often illiteracy (48.4%). Compared with the non-PMOP group, women with PMOP had a higher mean age (PMOP vs. non-PMOP: 64.6 ± 8.3 vs. 59.8 ± 7.8), and had higher average numbers of pregnancies (3.8 ± 2.0 vs. 3.4 ± 1.7) and livebirths (2.7 ± 1.6 vs. 2.3 ± 1.3). The total PA (17.6 ± 15.6 METs-h/day vs. 24.0 ± 18.0 METs-h/day), and PA through occupational (6.8 ± 12.9 MET-h/day vs. 12.4 ± 16.1 MET-h/day), transportation (0.9 ± 1.9 MET-h/day vs. 1.4 ± 2.3 MET-h/day), housework (7.3 ± 6.4 MET-h/day vs. 8.0 ± 6.8 MET-h/day) were lower in participants with PMOP than in those without PMOP (all P value < 0.05) (Table 1).
Table 1.
Basic characteristics of study participants
| Variable | Percentage (%) or Mean ± SD | P value | ||
|---|---|---|---|---|
| Total (n = 27,459) |
PMOP | |||
| No (n = 25,920) | Yes (n = 1,539) | |||
| Demographics | ||||
| Age | 60.1 ± 7.9 | 59.8 ± 7.8 | 64.6 ± 8.3 | < 0.001 |
| Residential location | ||||
| Urban | 13,083 (47.7) | 12,341 (47.7) | 742 (48.2) | 0.710 |
| Rural | 14,325 (52.3) | 13,528 (52.3) | 797 (51.8) | |
| Ethnicity | ||||
| Han | 14,701 (53.5) | 13,752 (53.1) | 949 (61.7) | 0.012 |
| Minority | 12,758 (46.5) | 12,168 (46.9) | 590 (38.3) | |
| Annual household income (RMB) | ||||
| < 12,000 | 6,512 (23.7) | 6,131 (23.7) | 381 (24.8) | 0.140 |
| 12,000–19,999 | 5,754 (21.0) | 5,429 (21.0) | 325 (21.1) | |
| 20,000–59,999 | 9,475 (34.5) | 8,976 (34.7) | 499 (32.4) | |
| ≥ 60,000 | 5,694 (20.7) | 5,360 (20.6) | 334 (21.7) | |
| Working environment | ||||
| Outdoor | 11,509 (41.9) | 11,094 (42.8) | 415 (27.0) | < 0.001 |
| Indoor | 9,460 (34.5) | 8,776 (33.9) | 684 (44.4) | |
| Others | 6,490 (23.6) | 6,050 (23.3) | 440 (28.6) | |
| Education level | ||||
| Illiteracy | 12,520 (45.6) | 11,775 (45.4) | 745 (48.4) | 0.007 |
| Primary school | 6,986 (25.4) | 6,592 (25.4) | 394 (25.6) | |
| Junior high school | 4,912 (17.9) | 4,683 (18.1) | 229 (14.9) | |
| High school | 2,256 (8.2) | 2,120 (8.2) | 136 (8.8) | |
| Junior college or higher | 784 (2.9) | 749 (2.9) | 35 (2.3) | |
| Physical activity | ||||
| Total PA (METs-h/day) | 23.6 ± 17.9 | 24.0 ± 18.0 | 17.6 ± 15.6 | < 0.001 |
| The level of total PA | ||||
| Low | 13,668 (50.0) | 12,640 (49.0) | 1,028 (67.1) | < 0.001 |
| High | 13,668 (50.0) | 13,164 (51.0) | 504 (32.9) | |
| Occupational PA (MET-h/day) | 12.1 ± 16.0 | 12.4 ± 16.1 | 6.8 ± 12.9 | |
| The level of occupational PA | ||||
| Low | 13,794 (50.5) | 12,733 (49.3) | 1,061 (69.3) | < 0.001 |
| High | 13,542 (49.5) | 13,071 (50.7) | 471 (30.7) | |
| Transportation PA (MET-h/day) | 1.4 ± 2.2 | 1.4 ± 2.3 | 0.9 ± 1.9 | < 0.001 |
| The level of transportation PA | ||||
| Low | 13,724 (50.2) | 12,742 (49.4) | 982 (64.1) | < 0.001 |
| High | 13,612 (49.8) | 13,062 (50.6) | 550 (35.9) | |
| Housework PA (MET-h/day) | 8.0 ± 6.8 | 8.0 ± 6.8 | 7.3 ± 6.4 | < 0.001 |
| The level of housework PA | ||||
| Low | 14,796 (54.1) | 13,914 (53.9) | 882 (57.6) | < 0.001 |
| High | 12,540 (45.9) | 11,890 (46.1) | 650 (42.4) | |
| Leisure PA (MET-h/day) | 2.2 ± 3.5 | 2.2 ± 3.5 | 2.7 ± 3.9 | < 0.001 |
| The level of leisure PA | ||||
| Low | 15,169 (55.6) | 14,403 (55.8) | 766 (50.0) | < 0.001 |
| High | 12,167 (44.4) | 11,401 (44.2) | 766 (50.0) | |
| Pregnancy and livebirth | ||||
| Reproductive status | ||||
| Age at menopause | 48.3 ± 4.5 | 48.3 ± 4.4 | 48.2 ± 4.8 | 0.440 |
| Age at menarche | 15.5 ± 2.3 | 15.5 ± 2.3 | 15.8 ± 2.3 | < 0.001 |
| Pregnancies | 3.4 ± 1.7 | 3.4 ± 1.7 | 3.8 ± 2.0 | < 0.001 |
| Number of pregnancies | ||||
| 0 | 287 (1.0) | 257 (1.0) | 30 (2.0) | < 0.001 |
| 1–2 | 8,651 (31.6) | 8,254 (31.9) | 397 (25.9) | |
| 3–4 | 12,757 (46.6) | 12,112 (46.8) | 645 (42.2) | |
| ≥ 5 | 5,708 (20.8) | 5,250 (20.3) | 458 (29.9) | |
| Livebirths | 2.3 ± 1.3 | 2.3 ± 1.3 | 2.7 ± 1.6 | < 0.001 |
| Number of livebirths | ||||
| 0 | 538 (2.0) | 495 (1.9) | 43 (2.8) | < 0.001 |
| 1–2 | 17,800 (65.2) | 17,000 (66.0) | 800 (52.5) | |
| 3–4 | 7,365 (27.0) | 6,864 (26.6) | 501 (32.9) | |
| ≥ 5 | 1,587 (5.8) | 1,407 (5.5) | 180 (11.8) | |
| History of contraceptives use | ||||
| No | 25,377 (92.4) | 23,925 (92.3) | 1,452 (94.3) | 0.003 |
| Yes | 2,082 (7.6) | 1,995 (7.7) | 87 (5.7) | |
| History of hysterectomy or oophorectomy | ||||
| No | 25,031 (91.2) | 23,622 (91.1) | 1,409 (91.6) | 0.610 |
| Yes | 2,427 (8.8) | 2,297 (8.9) | 130 (8.4) | |
| Total breastfeeding duration (months) | 32.2 ± 26.1 | 31.9 ± 25.8 | 37.4 ± 31.2 | < 0.001 |
| Lifestyle | ||||
| Smoking status | ||||
| No | 27,136 (98.82) | 25,620 (98.8) | 1,516 (98.5) | 0.177 |
| Yes | 258 (0.94) | 58 (0.2) | 7 (0.5) | |
| Quit | 65(0.24) | 242(1.0) | 16(1.0) | |
| Vitamin D or calcium supplements | ||||
| No | 22,198 (80.8) | 20,952 (80.8) | 1,246 (81.0) | 0.920 |
| Yes | 5,261 (19.2) | 4,968 (19.2) | 293 (19.0) | |
| AMED | 24.7 ± 4.4 | 24.7 ± 4.4 | 24.7 ± 4.6 | 0.860 |
| Calcium intake (mg/day) | 379.5 ± 195.7 | 378.9 ± 194.8 | 388.2 ± 210.5 | 0.219 |
| Physical measurement variable | ||||
| BMI (kg/m2) | 24.2 ± 3.7 | 24.2 ± 3.6 | 24.8 ± 4.3 | < 0.001 |
| Ultrasonic BMD (mg/cm3) | 4022.3 ± 185.9 | 4025.5 ± 169.3 | 3962.7 ± 377.1 | < 0.001 |
| QUI | 65.7 ± 29.4 | 66.2 ± 26.8 | 56.3 ± 59.6 | < 0.001 |
| Environment variables | ||||
| Ultraviolet radiation B (KJ/m2) | 59,702.2 ± 25,686.4 | 59,266.5 ± 25,401.0 | 67,033.5 ± 29,133.8 | < 0.001 |
| Temperature (℃) | 15.9 ± 3.5 | 16.0 ± 3.4 | 15.1 ± 5.1 | < 0.001 |
| Humidity (%) | 72.6 ± 9.4 | 72.6 ± 9.4 | 72.5 ± 10.0 | < 0.001 |
AMED Alternate mediterranean diet, BMD Bone mineral density, BMI Body mass index, kg/m2 Kilogram per square meter, KJ/m2 Kilojoule per square meter, MET Metabolic equivalent, mg/cm3 Milligram per cubic centimeter, PA Physical activity, PMOP Postmenopausal osteoporosis, QUI Quantitative ultrasound index, SD Standard deviation
Associations of the number of pregnancies and livebirths with the odds of postmenopausal osteoporosis
After adjusting for covariates, the number of pregnancies and livebirths were significantly associated with a higher odd for PMOP. Each increase in the number of pregnancies elevated 1.05-fold odds for PMOP (OR [95% CI]: 1.05 [1.02, 1.09]), and each increase in the number of livebirths elevated 1.21-fold odds for PMOP (OR [95% CI]: 1.21 [1.15, 1.28]). Additionally, both the number of pregnancies and livebirths were negatively correlated with ultrasonic BMD and QUI (Table 2).
Table 2.
The associations of the number of pregnancies and livebirths with postmenopausal osteoporosis
| OR (95%CI) | β (95%CI) | |||||
|---|---|---|---|---|---|---|
| Postmenopausal osteoporosis | Ultrasonic BMD | QUI | ||||
| Model 1 | Model 2 | Model 1 | Model 2 | Model 1 | Model 2 | |
| Pregnancies | 1.12(1.09, 1.15)*** | 1.05(1.02, 1.09)** | -4.02 (-5.34, -2.71)*** | -0.56 (-2.06, 0.94) | -0.64 (-0.84, -0.43)*** | -0.09 (-0.33, 0.15) |
| Number of pregnancies | ||||||
| 0 | 2.43(1.64, 3.59)*** | 2.56 (1.70, 3.84)*** | -0.13 (-22.90, 22.64) | 3.80 (-18.98, 26.58) | -0.02 (-3.62, 3.58) | 0.60 (-3.00, 4.20) |
| 1–2 | Ref | Ref | Ref | Ref | Ref | Ref |
| 3–4 | 1.11 (0.97, 1.26) | 0.95 (0.83, 1.09) | -7.31 (-12.61, -2.00)** | 1.14 (-4.33, 6.60) | -1.16 (-2.00, -0.32)** | 0.18 (-0.69, 1.04) |
| ≥ 5 | 1.81 (1.58, 2.08)*** | 1.26 (1.07, 1.48) | -21.37 (-27.83, -14.90)*** | -4.10 (-11.32, 3.10) | -3.38 (-4.4, -2.36)*** | -0.65(-1.79, 0.49) |
| Livebirths | 1.22 (1.18, 1.26)*** | 1.21 (1.15, 1.28)*** | -10.01 (-11.72, -8.29)*** | -10.01 (-11.72, -8.29)*** | -1.58 (-1.85, -1.31) | -1.58 (-1.85, -1.31)*** |
| Number of livebirths | ||||||
| 0 | 1.85 (1.34, 2.54)*** | 2.63 (1.76, 3.92)*** | 0.11 (-16.09, 16.31) | 4.33 (-17.80, 26.46) | 0.02 (-2.55, 2.58) | 0.69 (-2.81, 4.18) |
| 1–2 | Ref | Ref | Ref | Ref | Ref | Ref |
| 3–4 | 1.55 (1.38, 1.74)*** | 1.38 (1.19, 1.59)*** | -22.22 (-27.32, -17.13)*** | -6.01 (-12.39, 0.38) | -3.51 (-4.32, -2.71)*** | -0.95 (-1.96, 0.06) |
| ≥ 5 | 2.72 (2.29, 3.22)*** | 2.18(1.72, 2.77)*** | -41.04 (-50.55, -31.52)*** | -18.90 (-30.98, -6.82)** | -6.49 (-7.99, -4.99)*** | -2.99 (-4.90, -1.08)** |
Model 1: Adjusted for independent variables
Model 2: Adjusted covariates include age, annual household income, education level, ethnicity, residential location, working environment, smoking status, AMED, vitamin D or calcium supplements, calcium intake, age at menarche, age at menopause, total breastfeeding duration, history of contraceptive use, and history of hysterectomy or oophorectomy
AMED Alternate mediterranean diet, BMD Bone mineral density, CI Confidence interval, OR Odds ratios, QUI Quantitative ultrasound index, Ref Reference
*P < 0.05, **P < 0.01, ***P < 0.001
The modifying effects of PA
All types of PA were associated with more favorable bone health, and higher levels of PA tended to show stronger associations (Table S2). In exploratory interaction analyses, housework PA showed a potential interaction with the number of livebirths in relation to the odds of PMOP (P for interaction < 0.05, q < 0.05). Specifically, high levels of housework PA attenuated the adverse association of an increased number of livebirths on PMOP odds (Fig. 1a). Regarding BMD and QUI, total PA, occupational PA, and transportation PA modified the relationships between the number of pregnancies and livebirths and lower BMD and QUI (all P for interaction < 0.05, q < 0.05; Fig. 1b and c).
Fig. 1.
The modifying effects of PA on the association of the number of pregnancies and livebirths with postmenopausal osteoporosis, ultrasonic BMD and QUI. a The modifying effects of PA on the association of the number of pregnacies and livebirths with postmenopausal osteoporosis; (b) The modifying effects of PA on the association of the number of pregnacies and livebirths with ultrasonic BMD; (c) The modifying effects of PA on the association of the number of pregnacies and livebirths with QUI. BMD: Bone mineral density; CI: Confidence interval; OR: Odds ratios; PA: Physical activity; QUI: Quantitative ultrasound index. *P < 0.05, **P < 0.01, ***P < 0.001
Stratified and sensitivity analysis
In stratified analyses, associations of the number of pregnancies and livebirths with poorer bone health were observed among postmenopausal women aged > 60 years, with age at menopause ≤ 49 years, age at menarche ≥ 15 years, and a total breastfeeding duration ≥ 24 months (Tables S3). The total PA modified the associations between the number of livebirths and PMOP odds among age, age at menopause, age at menarche and a total breastfeeding duration of ≥ 24 months. A similar pattern was observed for the association between the number of pregnancies and PMOP odds among women aged > 60 years, with menopause ≤ 49 years, menarche ≥ 15 years, and total breastfeeding duration ≥ 24 months. Within these strata, women with higher levels of PA tended to show more favorable bone health indicators compared with those with lower levels of PA.
All the associations remained robust after excluding women with previous fractures or no pregnancies and after adjusting for ultraviolet light intensity, temperature, and humidity. Consistent findings were also observed when BUA and SOS were treated as continuous outcomes. However, the modifying effects were not consistently observed in the sensitivity analyses. (Tables S4-S11).
Discussion
To our knowledge, this is the first study to explore the potential modifying role of PA in the associations of the number of pregnancies and livebirths with the odds of PMOP. This study demonstrated that the number of pregnancies and livebirths were associated with bone health, and these associations were modified by the total PA and specific PA domains. High levels of total PA, particularly occupational and transportation PA, attenuated the adverse associations of the number of pregnancies and livebirths with bone health.
Pregnancy is associated with maternal bone loss. During this period, the calcium requirements of the fetus may lead to a reduction in maternal bone mineral density. Maternal calcium supply was essential for fetal skeletal development, with the physiological requirement of a full-term singleton fetus being approximately 30 g [8, 30]. This process was mediated through the increased secretion of parathyroid hormone and enhanced glomerular calcium filtration during pregnancy [31]. Additionally, the odds of PMOP may be associated with the number of pregnancies and livebirths. Previous studies have found that increased number of livebirths was associated with an elevated odds of lumbar osteoporotic fractures in postmenopausal women, and each additional livebirth was associated with a 13% increased odds of lumbar spine osteoporosis [27, 32]. Paradoxically, women with no pregnancy or no livebirth were associated with higher odds of PMOP. A drop in serum estradiol levels following human chorionic gonadotropin administration was linked to lower clinical pregnancy and livebirth rates in IVF cycles, and low estrogen levels were associated with higher odds of PMOP [33, 34]. Furthermore, nulliparous postmenopausal women typically had lower fertility and estrogen levels during their menstrual cycles, which may be associated with higher odds of PMOP [35].
Domain -specific types of PA may benefit bone health. As a major source of moderate-to-vigorous PA, occupational PA often involves walking or standing for extended periods, occupational lifting [36, 37]. Such mechanical loads could contribute bone strength through enhance bone mass and the structural skeletal adaptation [38]. A similar study have found that lifelong agricultural PA had a positive impact on postmenopausal women’s bone health [39]. Correspondingly, housework PA was negatively associated with PMOP odds. One study in American older women contended that heavy housework, such as vacuuming and mopping, benefit their bone [40]. Even though transportation PA and leisure PA accounted for a relatively small proportion of total PA, both modalities nonetheless conferred a significant reduction in the odds for PMOP [41, 42].
Another interesting finding of this study was the high levels of housework PA have been shown to modify the associations between the number of livebirths and PMOP odds. This may be attributed to the mechanical stimulation of the musculoskeletal system induced by PA. Such stimulation facilitated endocrine or paracrine crosstalk between bone and muscle by actin released from muscle at the periosteal interface or in the blood stream, such as irisin, thereby increasing bone accumulation [43, 44]. Moreover, the interaction of mechanical loading, hormones or cytokines, and signaling pathways induced by exercise may increase bone formation while decreasing bone resorption, which may be associated with more favorable skeletal health and prevent osteoporosis [45]. Thus, high levels of housework PA may attenuate the adverse association between livebirths and PMOP.
Associations of the number of pregnancies and livebirths with ultrasound BMD and QUI have been found to be modified by domain-specific PA. Previous studies have demonstrated that resistance training could increase BMD levels at 1-year postpartum, and combined exercise interventions were also effective in preserving BMD in postmenopausal women across the entire body [46, 47]. Furthermore, in older postmenopausal women, exercise training may enhance BMD, particularly in the femoral neck, lumbar spine, and trochanter regions [48]. This study observed that maintaining high levels of PA and engaging in specific types of PA were positively associated with bone health in postmenopausal women, and may attenuate the associations between bone health and the number of pregnancies and livebirths.
Our study suggested that age and certain reproductive factors were associated with less favorable bone health status in postmenopausal women. For example, advanced age was an adverse factor for PMOP odds. Previous study have found that the BMD peaked around age 35 and then declined with advancing age, with a significant acceleration after menopause [49]. The temporal dynamics of estrogen exposure critically modulate bone health across a woman’s lifespan. Earlier menopause and later menarche were linked to limited estrogen exposure, which was significantly associated with increased odds of fractures at all sites [4, 50, 51]. Besides, with the self-limiting bone loss and rapid decline in BMD during menopause, women who experienced menopause later in life tended to have better skeletal health compared with their peers [4, 52]. Prolonged breastfeeding duration was associated with poorer bone health, and may synergistically exacerbate bone health impairment with parity. During lactation, calcium loss through breast milk averaged 300–400 mg per day, and women who breastfed for 24 months or more had significantly higher odds of lumbar spine and femoral neck osteoporosis [9].
Several limitations should be mentioned in this study. First, this is a cross-sectional study. Although we observed significant associations, we cannot determine the temporal sequence between PA and PMOP, and reverse causation cannot be ruled out. The interaction findings in this study were exploratory, and conducting subgroup analysis with limited sample size would reduce statistical power. Second, using self-reported questionnaires to collect data on reproductive factors and PA may introduce recall bias and increase the risk of misclassifying PA intensity, particularly in occupational and housework domains. Further studies are warranted to utilize wearable devices to capture the detailed PA types. Third, we did not use dual-energy X-ray absorptiometry to assess the BMD but instead used quantitative ultrasound bone densitometer-based T-score as an alternative, which may affect the accuracy of measurement and potential misclassification of osteoporosis status. However, in large sample size population survey, quantitative ultrasound bone densitometer could be more feasible and convenient [53]. Fourth, excluding participants with missing data may have introduced selection bias, although the proportion excluded was relatively small. Fifth, the generalizability of our findings may be limited to Chinese postmenopausal women, and caution is needed when extrapolating to other countries.
Conclusion
An increased number of pregnancies and livebirths was positively associated with the odds of PMOP. However, high levels of total and domain-specific PA were associated with lower odds of PMOP, and high levels of housework PA attenuated the adverse association of an increased number of livebirths on PMOP odds. Postmenopausal women with lower PA levels, older age, earlier menopause, later menarche, and longer breastfeeding duration exhibited particularly high odds of PMOP. These findings suggest that PA may play an important role in reducing the odds of PMOP, highlighting its potential as a modifiable factor for bone health in this population.
Supplementary Information
Acknowledgements
The authors would like to thank all consenting participants for their cooperation.
Authors’ contributions
Xiru Wang was responsible for data curation, conceptualization, methodology, software, formal analysis, validation, writing – original draft. Bin Yu was responsible for software, validation, writing – original draft. Jie Xu, Jing Huang, Delin Han and Yanjiao Wang were involved in investigation. Peng Jia was responsible for writing – original draft. Shujuan Yang was responsible for Data curation, Conceptualization, Methodology, Formal analysis, Validation, Writing – original draft, Writing – review & editing.
Funding
This study was supported by the National Key R&D Program of China (2023YFC3604702), National Natural Science Foundation of China (723B2017), Chongqing Maternal and Child Health Hospital (CQFYJB01005), Science and Technology Major Project of Tibetan Autonomous Region of China (XZ202201ZD0001G), and Science and technology project of Tibet Autonomous Region (XZ202303ZY0007G).
Data availability
Data will be made available to the public after publication of study findings upon request from the corresponding author.
Declarations
Ethics approval and consent to participate
Ethical approval for this study was obtained from the Sichuan University Medical Ethical Review Board (K2016038). All individuals voluntarily participated in this study and obtained informed consent. This study was conducted in accordance with the Declaration of Helsinki.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Xiru Wang, Bin Yu and Jie Xu contributed equally to this work.
References
- 1.Rentzeperi E, Pegiou S, Tsakiridis I, Kalogiannidis I, Kourtis A, Mamopoulos A, Athanasiadis A, Dagklis T. Diagnosis and Management of Osteoporosis: A Comprehensive Review of Guidelines. Obstet Gynecol Surv. 2023;78(11):657–681. [DOI] [PubMed]
- 2.Lee CNY, Lam SC, Tsang AYK, Ng BTY, Leung JCY, Chong ACY. Preliminary investigation on prevalence of osteoporosis and osteopenia: Should we tune our focus on healthy adults? Japan J Nurs Sci. 2015;12(3):232–48. [DOI] [PubMed] [Google Scholar]
- 3.Reginster J-Y, Burlet N. Osteoporosis: a still increasing prevalence. Bone. 2006;38(2 Suppl 1):S4–9. [DOI] [PubMed] [Google Scholar]
- 4.Chen J, Liang X, Wang Y, Dejiquzong, Zhang Y, Chen L, Liu Q, Zhao X. The association between age at menopause and bone health in Southwest China women: mediation effect of body mass index. BMC Public Health. 2024;24(1):3153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pandey MK, Gupta SC, Karelia D, Gilhooley PJ, Shakibaei M, Aggarwal BB. Dietary nutraceuticals as backbone for bone health. Biotechnol Adv. 2018;36(6):1633–48. [DOI] [PubMed] [Google Scholar]
- 6.Li J, Chen X, Lu L, Yu X. The relationship between bone marrow adipose tissue and bone metabolism in postmenopausal osteoporosis. Cytokine Growth Factor Rev. 2020;52:88–98. [DOI] [PubMed] [Google Scholar]
- 7.Sanz-Salvador L, García-Pérez MÁ, Tarín JJ, Cano A. ENDOCRINOLOGY IN PREGNANCY: Bone metabolic changes during pregnancy: a period of vulnerability to osteoporosis and fracture. Eur J Endocrinol. 2015;172(2):R53–65. [DOI] [PubMed] [Google Scholar]
- 8.Pitkin RM. Calcium metabolism in pregnancy and the perinatal period: a review. Am J Obstet Gynecol. 1985;151(1):99–109. [DOI] [PubMed] [Google Scholar]
- 9.Capozzi A, Scambia G, Lello S. Bone metabolism in pregnancy and lactation. Minerva Obstet Gynecol. 2021;73(6):697–703. [DOI] [PubMed] [Google Scholar]
- 10.Kalkwarf HJ, Specker BL. Bone mineral changes during pregnancy and lactation. Endocrine. 2002;17(1):49–53. [DOI] [PubMed] [Google Scholar]
- 11.Gur A, Nas K, Cevik R, Sarac AJ, Ataoglu S, Karakoc M. Influence of number of pregnancies on bone mineral density in postmenopausal women of different age groups. J Bone Miner Metab. 2003;21(4):234–41. [DOI] [PubMed] [Google Scholar]
- 12.Cure-Cure C, Cure-Ramírez P, Terán E, López-Jaramillo P. Bone-mass peak in multiparity and reduced risk of bone-fractures in menopause. Int J Gynaecol Obstet. 2002;76(3):285–91. [DOI] [PubMed] [Google Scholar]
- 13.Grainge MJ, Coupland CA, Cliffe SJ, Chilvers CE, Hosking DJ. Reproductive, menstrual and menopausal factors: which are associated with bone mineral density in early postmenopausal women? Osteoporos international: J established as result cooperation between Eur Foundation Osteoporos Natl Osteoporos Foundation USA. 2001;12(9):777–87. [DOI] [PubMed] [Google Scholar]
- 14.Colaianni G, Cinti S, Colucci S, Grano M. Irisin and musculoskeletal health. Ann N Y Acad Sci. 2017;1402(1):5–9. [DOI] [PubMed] [Google Scholar]
- 15.Jang S-N, Choi Y-H, Choi M-G, Kang S-H, Jeong J-Y, Choi Y-J, Kim D-H. [Prevalence and associated factors of osteoporosis among postmenopausal women in Chuncheon: Hallym Aging Study (HAS)]. J Prev Med Public Health = Yebang Uihakhoe Chi. 2006;39(5):389–96. [PubMed] [Google Scholar]
- 16.Ji J, Hou Y, Li Z, Zhou Y, Xue H, Wen T, Yang T, Xue L, Tu Y, Ma T. Association between physical activity and bone mineral density in postmenopausal women: a cross-sectional study from the NHANES 2007–2018. J Orthop Surg Res. 2023;18(1):501. [DOI] [PMC free article] [PubMed]
- 17.Zhao X, Hong F, Yin J, Tang W, Zhang G, Liang X, Li J, Cui C, Li X. Cohort Profile: the China Multi-Ethnic Cohort (CMEC) study. Int J Epidemiol. 2021;50(3):721–l721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Liu DH, Lin CS, Wu PC. Osteoporosis self-assessment tool for Asians and calcaneal quantitative ultrasound for identifying primary osteoporosis in Taiwanese postmenopausal women. Front Endocrinol (Lausanne). 2025;16:1639176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Magkos F, Manios Y, Babaroutsi E, Sidossis LS. Quantitative ultrasound calcaneus measurements: normative data for the Greek population. Osteoporos international: J established as result cooperation between Eur Foundation Osteoporos Natl Osteoporos Foundation USA. 2005;16(3):280–8. [DOI] [PubMed] [Google Scholar]
- 20.Osteoporosis WoCGftD T of, Geriatrics S, OSoCAoGa, Yuanzheng M, Yipeng W, Qiang L, Chunlin L, Xun M, Yongjun W, Lianfu D et al. 2018 Chinese guideline for the diagnosis and treatment of senile osteoporosis. Chin J Health Manag. 2018;12(6):484–509.
- 21.McLeod KM, Johnson S, Rasali D, Verma A. Discriminatory Performance of the Calcaneal Quantitative Ultrasound and Osteoporosis Self-Assessment Tool to Select Older Women for Dual-Energy X-ray Absorptiometry. J Clin Densitom. 2015;18(2):157–64. [DOI] [PubMed] [Google Scholar]
- 22.Zhang Y, Chen C, Wu S, Nie C, Hu Y, Zhong J, Hong F. Analysis of the association between mixed exposure to multiple metals and comorbidity of osteopenia or osteoporosis: baseline data from the Chinese Multi-Ethnic Cohort study (CMEC). BMC Public Health. 2025;25(1):680. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Yang L, Hu X, Wu W, Pan W, Chen S, Lin X, Gu X. The inverse association between parity and bone health is independent of lifestyle in postmenopausal Chinese women. Endocr J. 2020;67(10):1039–45. [DOI] [PubMed] [Google Scholar]
- 24.Lenora J, Lekamwasam S, Karlsson MK. Effects of multiparity and prolonged breast-feeding on maternal bone mineral density: a community-based cross-sectional study. BMC Womens Health. 2009;9:19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett DR, Tudor-Locke C, Greer JL, Vezina J, Whitt-Glover MC, Leon AS. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43(8):1575–81. [DOI] [PubMed] [Google Scholar]
- 26.Du H, Bennett D, Li L, Whitlock G, Guo Y, Collins R, Chen J, Bian Z, Hong LS, Feng S, et al. Physical activity and sedentary leisure time and their associations with BMI, waist circumference, and percentage body fat in 0.5 million adults: the China Kadoorie Biobank study. Am J Clin Nutr. 2013;97(3):487–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Heidari B, Hosseini R, Javadian Y, Bijani A, Sateri MH, Nouroddini HG. Factors affecting bone mineral density in postmenopausal women. Arch Osteoporos. 2015;10:15. [DOI] [PubMed] [Google Scholar]
- 28.Lee EN. Effects of Parity and Breastfeeding Duration on Bone Density in Postmenopausal Women. Asian Nurs Res (Korean Soc Nurs Sci). 2019;13(2):161–7. [DOI] [PubMed] [Google Scholar]
- 29.Movassagh EZ, Vatanparast H. Current Evidence on the Association of Dietary Patterns and Bone Health: A Scoping Review. Adv Nutr. 2017;8(1):1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.White CP. Calcium metabolism in pregnancy and lactation. Obstetric Med. 2009;2(1):2–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kovacs CS. Maternal Mineral and Bone Metabolism During Pregnancy, Lactation, and Post-Weaning Recovery. Physiol Rev. 2016;96(2):449–547. [DOI] [PubMed] [Google Scholar]
- 32.Yang Y, Wang S, Cong H. Association between parity and bone mineral density in postmenopausal women. BMC Womens Health. 2022;22(1):87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kondapalli LA, Molinaro TA, Sammel MD, Dokras A. A decrease in serum estradiol levels after human chorionic gonadotrophin administration predicts significantly lower clinical pregnancy and live birth rates in in vitro fertilization cycles. Hum Reprod (Oxford England). 2012;27(9):2690–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Patel S, Homaei A, Raju AB, Meher BR. Estrogen: The necessary evil for human health, and ways to tame it. Biomed pharmacotherapy = Biomedecine pharmacotherapie. 2018;102:403–11. [DOI] [PubMed] [Google Scholar]
- 35.Hillier TA, Rizzo JH, Pedula KL, Stone KL, Cauley JA, Bauer DC, Cummings SR. Nulliparity and fracture risk in older women: the study of osteoporotic fractures. J Bone Mineral Research: Official J Am Soc Bone Mineral Res. 2003;18(5):893–9. [DOI] [PubMed] [Google Scholar]
- 36.Walker-Bone K, D’Angelo S, Syddall HE, Palmer KT, Cooper C, Coggon D, Dennison EM. Exposure to heavy physical occupational activities during working life and bone mineral density at the hip at retirement age. Occup Environ Med. 2014;71(5):329–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Pope DP, Hunt IM, Birrell FN, Silman AJ, Macfarlane GJ. Hip pain onset in relation to cumulative workplace and leisure time mechanical load: a population based case-control study. Ann Rheum Dis. 2003;62(4):322–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Karlsson MK, Rosengren BE. Training and bone - from health to injury. Scand J Med Sci Sports. 2012;22(4):e15–23. [DOI] [PubMed] [Google Scholar]
- 39.Damilakis J, Perisinakis K, Kontakis G, Vagios E, Gourtsoyiannis N. Effect of lifetime occupational physical activity on indices of bone mineral status in healthy postmenopausal women. Calcif Tissue Int. 1999;64(2):112–6. [DOI] [PubMed] [Google Scholar]
- 40.Ilich JZ, Brownbill RA. Habitual and low-impact activities are associated with better bone outcomes and lower body fat in older women. Calcif Tissue Int. 2008;83(4):260–71. [DOI] [PubMed] [Google Scholar]
- 41.Armstrong ME, Lacombe J, Wotton CJ, Cairns BJ, Green J, Floud S, Beral V, Reeves GK. The Associations Between Seven Different Types of Physical Activity and the Incidence of Fracture at Seven Sites in Healthy Postmenopausal UK Women. J bone mineral research: official J Am Soc Bone Mineral Res. 2020;35(2):277–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Brooke-Wavell K, Jones PR, Hardman AE. Brisk walking reduces calcaneal bone loss in post-menopausal women. Clin Sci (London England: 1979). 1997;92(1):75–80. [DOI] [PubMed] [Google Scholar]
- 43.Santos L, Elliott-Sale KJ, Sale C. Exercise and bone health across the lifespan. Biogerontology. 2017;18(6):931–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Vita F, Gangemi S, Pioggia G, Trimarchi F, Di Mauro D. Physical Activity and Post-Transcriptional Regulation of Aging Decay: Modulation of Pathways in Postmenopausal Osteoporosis. Med (Kaunas Lithuania). 2022;58(6):767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Tong X, Chen X, Zhang S, Huang M, Shen X, Xu J, Zou J. The Effect of Exercise on the Prevention of Osteoporosis and Bone Angiogenesis. Biomed Res Int. 2019;2019:8171897. [DOI] [PMC free article] [PubMed]
- 46.Colleran HL, Hiatt A, Wideman L, Lovelady CA. The Effect of an Exercise Intervention During Early Lactation on Bone Mineral Density During the First Year Postpartum. J Phys Act Health. 2019;16(3):197–204. [DOI] [PubMed] [Google Scholar]
- 47.Zhao R, Zhang M, Zhang Q. The Effectiveness of Combined Exercise Interventions for Preventing Postmenopausal Bone Loss: A Systematic Review and Meta-analysis. J Orthop Sports Phys Ther. 2017;47(4):241–51. [DOI] [PubMed] [Google Scholar]
- 48.Hejazi K, Askari R, Hofmeister M. Effects of physical exercise on bone mineral density in older postmenopausal women: a systematic review and meta-analysis of randomized controlled trials. Arch Osteoporos. 2022;17(1):102. [DOI] [PubMed] [Google Scholar]
- 49.Lamichhane AP. Osteoporosis-an update. JNMA. 2005;44(158):60–6. [PubMed] [Google Scholar]
- 50.Eastell R. Role of oestrogen in the regulation of bone turnover at the menarche. J Endocrinol. 2005;185(2):223–34. [DOI] [PubMed] [Google Scholar]
- 51.Chang HK, Chang DG, Myong JP, Kim JH, Lee SJ, Lee YS, Lee HN, Lee KH, Park DC, Kim CJ, et al. Bone mineral density among Korean females aged 20–50 years: influence of age at menarche (The Korea National Health and Nutrition Examination Survey 2008–2011). Osteoporos international: J established as result cooperation between Eur Foundation Osteoporos Natl Osteoporos Foundation USA. 2017;28(7):2129–36. [DOI] [PubMed] [Google Scholar]
- 52.Greendale GA, Sowers M, Han W, Huang MH, Finkelstein JS, Crandall CJ, Lee JS, Karlamangla AS. Bone mineral density loss in relation to the final menstrual period in a multiethnic cohort: results from the Study of Women’s Health Across the Nation (SWAN). J bone mineral research: official J Am Soc Bone Mineral Res. 2012;27(1):111–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Shen Z, Yu C, Guo Y, Bian Z, Wei Y, Du H, Yang L, Chen Y, Gao Y, Zhang X, et al. Weight loss since early adulthood, later life risk of fracture hospitalizations, and bone mineral density: a prospective cohort study of 0.5 million Chinese adults. Arch Osteoporos. 2020;15(1):60. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data will be made available to the public after publication of study findings upon request from the corresponding author.