Association between sarcopenia, its components, and osteoporosis in the FRISBEE cohort

Abstract

Background & objectives

Bone and muscle diseases are both highly prevalent in aging adults, but results from previous studies examining the relationship between sarcopenia and its components with osteoporosis are inconsistent. This study aimed to investigate the associations between sarcopenia, its components, and osteoporosis in older women from the Fracture RISk Brussels Epidemiological Enquiry (FRISBEE) cohort. Additionally, it explored the relationship between sarcopenia components and trabecular bone score (TBS).

Methods

This study is based on cross-sectional data from the FRISBEE cohort, involving 3560 community-dwelling postmenopausal women initially included between 2007 and 2013. Consecutively included participants were reassessed 10 years after inclusion, with evaluations including body composition by dual-energy X-ray absorptiometry (DXA), a medical questionnaire, and a comprehensive geriatric assessment. The diagnostic criteria for sarcopenia were: low muscle mass (appendicular skeletal muscle index [ASMi] < 5.5 kg/m²), low muscle strength (< 16 kg), and low gait speed (≤ 0.8 m/s), according to the European Working Group on Sarcopenia in Older People (EWGSOP2). Osteoporosis was defined using the extended criteria proposed by the National Bone Health Alliance Working Group.

Results

Among the 500 women included, with a median age of 77.4 years (74.7–81.8), 178 (35.6%) were diagnosed with osteoporosis. Significant correlations were shown between sarcopenia, its components, and bone mineral density (BMD), as well as with TBS. The strongest correlations were between handgrip strength and distal forearm BMD (r = 0.27; p < 0.001), and between ASMi and total hip BMD (r = 0.36; p < 0.001). Significant differences were observed according to osteoporotic status: handgrip strength, Short Physical Performance Battery (SPPB) test, gait speed, and ASMi were significantly lower in participants with osteoporosis (p < 0.001). After adjustment for all covariates, handgrip strength and gait speed were still significantly associated with osteoporosis with an odds ratio of 0.92 (0.88–0.97) and 0.33 (0.11–0.96), respectively. A significant association was also observed between sarcopenia and osteoporosis with a crude odds ratio of 3.88 (1.97–7.85).

Conclusions

Sarcopenia and its components, particularly handgrip strength and gait speed were significantly associated with osteoporosis. TBS was also significantly associated with all sarcopenia components. These findings may be helpful when evaluating bone health and fracture risk in clinical practice.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-025-09146-1.

Background

The prevalence of osteoporosis continually increases as the population ages. Low bone mass and loss of bone microarchitecture result in reduced bone strength and increased fracture risk [1]. The operational definition of osteoporosis relies on the densitometric criteria of the World Health Organization (WHO) and is based on bone mineral density (BMD) values, i.e., a T-score ≤−2.5 standard deviations (SD) at the hip, femoral neck, or spine compared to the mean of a young population. However, at least 50% of fragility fractures occur in patients with higher BMD values, prompting the consideration of a more inclusive definition [2]. The National Bone Health Alliance Working Group proposed classifying as osteoporotic, older individuals with a T-score ≤−2.5 at the hip, femoral neck or spine; those with a history of hip fracture regardless of the BMD value; those with a prior fragility fracture and a BMD T-score lower than − 1; and those with a FRAX score ≥ 3% for hip fracture or ≥ 20% for a major osteoporotic fracture (MOF) [3]. This extended definition of osteoporosis should reduce its underdiagnosis and undertreatment, leading to a decrease in fracture risk in this silent disease. This is essential since osteoporotic fractures are known to result in significant morbidity and mortality, with a high individual, societal, and economic burden [1].

Sarcopenia is another prevalent disease in the aging population, where a progressive, age-related decrease in muscle mass and strength occurs with adverse outcomes such as falls, fractures, functional disability, and increased mortality [4]. At present, there is no universal consensus on the definition of sarcopenia. According to the most recent revised definition of the European Working Group on Sarcopenia in Older People (EWGSOP2), the diagnosis of sarcopenia is confirmed by the presence of low muscle strength and low muscle mass. When low physical performance is also present, sarcopenia is considered “severe” [5].

There is conflicting evidence in the literature regarding the relationships between sarcopenia, its components, and osteoporosis [6–8]. Furthermore, few studies investigating these relationships have considered all three components of sarcopenia as defined by EWGSOP2, and the findings remain contradictory. Tiftik et al. reported that low handgrip strength was associated with a 1.6-fold greater risk of osteoporosis according to the WHO criteria [6]. Kirk et al. showed an association only between hip BMD and gait speed when evaluating all three components of sarcopenia [7]. In contrast, Elkhaheem et al. did not find any associations between grip strength, gait speed, or chair rise time and hip BMD [8]. Thus, there is still a lack of evidence proving that targeting sarcopenia could be a strategy for detecting and reducing fracture risk.

The objective of the study was to investigate the relationships between sarcopenia, its components (muscle strength, muscle mass, and physical performance), and bone mineral density and quality in community-dwelling older women from the Fracture RISk Brussels Epidemiological Enquiry (FRISBEE) study. The overall goal of the FRISBEE project is to validate and integrate several independent clinical risk factors (CRFs), such as sarcopenia and its components, for the development of fracture risk models.

Methods

Study design and participants

This study is based on cross-sectional data obtained from the ongoing FRISBEE prospective cohort study. The FRISBEE cohort comprises 3560 postmenopausal women who have been surveyed annually by telephone interview since their inclusion between 2007 and 2013. Baseline characteristics were collected by trained nurses during a face-to-face interview, and dual-energy X-ray absorptiometry (DXA) was performed on the same day. A 5-year fracture prediction model has been published [9] and validated [10]. A first prediction model for imminent fractures (within 2 years) after a fragility fracture has also been developed and published [11].

Of the 3560 women initially included, 2433 are still interviewed yearly. For this study, we invited all the women still followed annually and decided to include the first 500 respondents who were consecutively assessed. They were contacted to be re-evaluated with a medical questionnaire, a new DXA, and a comprehensive geriatric assessment. Written informed consent was obtained from each participant. The objective of this second part of the FRISBEE study is to assess the relationships between geriatric syndromes, osteoporosis, and fracture risk. The study protocol was approved by the Brugmann Ethics Committee (Number CE 2021/91). This report is based on the evaluation of the first 500 participants included in a consecutive manner.

Exclusion criteria

• Severe cognitive impairment (e.g. responses available only from family members).

• Nursing home residents.

• Severe comorbidities that will prevent DXA examinations and physical performance tests.

• Recent bedridden status (at least 3–4 weeks) due to a severe acute condition, which could impair DXA evaluation of body composition.

• Severe concomitant disease and short life expectancy (e.g. advanced cancer).

• Refusal to participate.

Demographic, clinical risk factors, and comprehensive geriatric assessment

Information was collected during the medical interview. The following characteristics included: birthdate, height and weight, daily alcohol and tobacco intake, weekly physical activity, number of falls in the past twelve months, medical history, and parental history of hip fracture, ongoing treatment, and particularly medications affecting BMD, including menopausal hormonal therapy (MHT), osteoporosis treatments, calcium and vitamin D supplements, and corticosteroids. The history of fractures was also recorded. Any reported fracture by the participant was verified by obtaining written radiological and/or surgical reports. Fractures not reported by study participants but found in their medical files and validated by radiological reports were also registered [12]. Causes of secondary osteoporosis, such as rheumatoid arthritis or early non-substituted menopause, were also listed. The comprehensive geriatric assessment was completed as follows: cognitive function was evaluated by the Mini-Mental State Evaluation (MMSE) and scored on 30 points; frailty was assessed by the Fried criteria (which include involuntary weight loss >4.5 kg in 12 months, exhaustion, low physical activity (<1 h/week of low-intensity physical activity), 4-meter walking time ≥6 s, and low handgrip strength value (cutoffs defined by taking BMI into consideration) [13].

Bone mineral density and body composition assessment

BMD of the lumbar spine, femoral neck, total hip, and non-dominant forearm was measured by DXA (Hologic, USA). Osteopenia was classically defined by a T-score between − 1 and − 2.5 at the hip, femoral neck, or lumbar spine. The diagnosis of osteoporosis in our study was established based on the diagnostic criteria proposed by the National Bone Health Alliance Working Group, i.e., a T-score ≤−2.5 at the lumbar spine, total hip or femoral neck; a history of non-traumatic hip fracture regardless of the BMD; or a history of low-trauma vertebral, pelvis, proximal humerus, or distal forearm fracture in participants with osteopenia. A FRAX prediction score ≥ 3% for the hip or ≥ 20% for a MOF was also considered a criterion for a diagnosis of osteoporosis [3].

Body composition, including total-body bone mineral (TBBM), whole-body fat (WBF), total BMD, and appendicular skeletal mass (ASM), was assessed using whole-body DXA (Hologic, Horizon). The trabecular bone score (TBS), which reflects the trabecular bone microarchitecture, was also determined. A low TBS (< 1.20) indicates a degraded microarchitecture and is associated with a greater risk of fracture [14]. Subjects with an extreme body mass index (BMI) (< 15 kg/m² or > 37 kg/m²) were not considered for TBS measurement.

Sarcopenia and its components

We screened for sarcopenia using the SARC-F questionnaire. It consists of five questions and a score ≥ 4 suggests “probable sarcopenia” [15]. Handgrip strength (kg) was assessed using a JAMAR© hydraulic dynamometer (USA). We measured the strength three times in both hands, and the highest value was used for subsequent analysis, following standardized procedures (Southampton protocol) [16]. We used the cutoff of < 16 kg to define “probable sarcopenia” according to the revised consensus on definition and diagnosis of sarcopenia (EWGSOP2) [5]. We evaluated the physical performance of the lower limbs with the Short Physical Performance Battery (SPPB). The scoring ranges from 0 to 12 and is based on three different tests: the ability to stand for 10 s with feet in three different positions, two timed trials of a 4-meter walk, and the time to stand up from a chair five times in a row. This score has been specifically developed, validated, and used in community-dwelling older individuals [17], which corresponds to the study population. According to the EWGSOP2 definition, the measurement of appendicular skeletal mass was adjusted for body size using height squared (ASMi, in kg/m²) with the cutoff of <5.5 kg/m² used to confirm sarcopenia. The gait speed (m/s) evaluated in the SPPB testing pointed to a risk for severe sarcopenia when ≤0.8 m/s [5].

Statistical analyses

Statistical analyses were performed using Stata-SE 17.0 (StataCorp, College Station, Texas, USA), graphics were made with R (4.4.1) and its packages ggplot and GGally [18]. Descriptive statistics were expressed as means ± SD or median (P25-P75) for continuous variables, and as absolute (n) or relative (%) frequencies for categorical variables. After checking for the normality of distributions and homoscedasticity, Pearson or Spearman correlation coefficients were computed to look for and estimate the strength of associations between sarcopenia components and the lumbar spine, femoral neck, total hip, distal forearm, whole-body BMD, and TBS.

Distributions of the participants’ characteristics were compared between osteoporotic and non-osteoporotic groups: t-tests, Kruskal-Wallis, and Pearson’s Chi² were performed according to the variable characteristics.

The associations of sarcopenia components and osteoporotic status were estimated using odds ratios (OR) obtained through logistic regression. First, univariate regression models were performed. Risk factors associated with osteoporotic status (p < 0.3) were subsequently included in a multivariate regression model including the EWGSOP2 variable. A stepwise backward selection procedure was applied to identify significant covariates, which were then used in multivariate regression analyses of the individual components of sarcopenia.

Models’ adjustments were evaluated with Hosmer and Lemeshow goodness-of-fit test and their specification with link test (Stata-SE). ORs were reported with 95% confidence intervals (95% CIs). Collinearity diagnosis was checked with variance inflation factors (VIF) [19]. Influential points were identified with delta-deviance (Δd), which represents the impact of deleting the observation on the model deviance statistics. Observations whose Δd was >4 were deleted [20].

Mean or median handgrip strength, SPPB score, gait speed, and ASMi were compared between participants with and without osteoporosis, as defined previously, by using t-tests or Kruskal-Wallis tests. All statistical tests were two-tailed and p < 0.05 was considered statistically significant.

Results

Participants’ characteristics

Our cohort of 500 community-dwelling postmenopausal women had a median age of 77.4 years (range: 74.7–81.8) and a median BMI of 26.2 kg/m² (range: 23.0–29.9). The characteristics, physical test results, and BMD measurements of the study participants are described in Table 1S (See Supplementary material, Table 1S).

A total of 178 women (35.6%) were diagnosed with osteoporosis based on the criteria of the National Bone Health Alliance Working Group. Osteoporosis defined only by a BMD T-score ≤−2.5 was present in 118 participants (23.6%), and osteopenia in 274 participants (54.8%). We observed low muscle strength (handgrip strength <16 kg) in 110 participants (22%), and 180 (36.1%) had a score ≤8 in the SPPB. The gait speed was slow (≤0.8 m/s) in 185 women (37.2%). The ASMi was low (< 5.5 kg/m²) in 183 participants (36.6%).

Potential risk factors associated with osteoporosis

As shown in Table 1, we have compared the prevalence of CRFs according to participants’ osteoporotic status, revealing significant differences between the two groups. Participants with osteoporosis were slightly older, had a lower BMI, had a more frequent history of early menopause, and were more likely to be receiving osteoporosis treatment, and were less likely to be receiving MHT.

Table 1.

Participants’ characteristics according to their osteoporosis status

Risk factors	n	Normal BMD and osteopenia* 322 (64,4%)	Osteoporosis* 178 (35.6%)	p-value
		Results**
Age (years)	500	76.8 (74.3–79.9)	79.2 (75.8–84.7)	< 0.001
BMI (kg/m²)	500	27.7 ± 5.0	25.0 ± 4.5	< 0.001
MMSE (/30)	485	28 (25–29)	28 (25–29)	0.71
Frailty status (≥ 3 Fried criteria)	499	15 (4.7%)	11 (6.2%)	0.47
SARC-F ≥ 4	499	23 (7.2%)	28 (15.7%)	0.002
Smoking	500	11 (3.4%)	14 (7.9%)	0.03
Alcohol (> 2 units/day)	500	20 (6.2%)	3 (1.7%)	0.02
History of parental hip fracture	488	41 (12.9%)	34 (19.9%)	0.04
Sedentary lifestyle	500	34 (10.6%)	20 (11.2%)	0.82
Rheumatoid arthritis	500	7 (2.2%)	2 (1.1%)	0.40
Early menopause	500	6 (1.9%)	14 (7.9%)	0.001
Secondary osteoporosis (early non-substituted menopause and rheumatoid arthritis excluded)	500	55 (17.1%)	39 (21.9%)	0.19
Calcium and/or Vitamin D supplements intake	499	280 (87.2%)	159 (89.3%)	0.49
Osteoporosis treatment	497	10 (3.1%)	18 (10.2%)	0.001
Menopausal hormone therapy	498	41 (12.8%)	9 (5.1%)	0.006
SSRI and/or PPI therapy	496	127 (39.8%)	65 (36.7%)	0.50
Corticosteroid therapy ≥ 3 months	498	12 (3.7%)	12 (6.8%)	0.13
Falls (≥ 2) in the past 12 months	499	40 (12.5%)	29 (16.3%)	0.24

BMD Bone Mineral Density, BMI Body Mass Index, MMSE Mini-Mental State Evaluation, SARC-F Strength, Ambulation, Rising from a chair, stair Climbing and history of Falling, SSRI Selective Serotonin Reuptake Inhibitor, PPI Proton Pump Inhibitor

*According to the National Bone Health Alliance (reference 3)

**Mean ± Standard Deviation or Medians with inter-quartile ranges, i.e. P50 (P25 - P75), or n (%)

Relationships between BMD at all sites, TBS, and sarcopenia components

We observed significant positive correlations between handgrip strength and ASMi with BMD at all sites, and between gait speed and total hip, femoral neck, and distal forearm BMD, as shown in Fig. 1. The strongest correlations were found with handgrip strength and ASMi, more specifically between handgrip strength and distal forearm BMD (r = 0.27; p<0.001), and between ASMi and total hip BMD (r = 0.35; p<0.001). Handgrip strength and gait speed, but not ASMi, were also significantly correlated with TBS, although the correlations were weak.

Fig. 1.

Scatterplots showing the correlation of EWGSOP2-defined sarcopenia components, bone mineral density (BMD) in the five measurement sites, and trabecular bone score (TBS). ASMi: Appendicular Sketelal Muscle index, ASM/height²; BMD: Bone Mineral Density; EWGSOP2: Revised European consensus on definition and diagnosis of sarcopenia or European Working Group on Sarcopenia in Older People 2 (Cruz-Jentoft et al., 2019); SPPB: Short Physical Performance Battery; TBS: Trabecular bone score, measured at vertebrae L1-L4. r = Pearson’s or Spearman’s correlation coefficient. * p<0.05; ** p<0.01; *** p<0.001. Each observation is represented as a light grey point. Dark grey or black indicate several overlying observations. Handgrip strength was expressed in kg, gait speed in m/s, ASMi in kg/m², BMD in g/cm². Spine BMD was measured at vertebrae L1-L4

Comparison of sarcopenia components according to osteoporosis status

Table 2 shows that all evaluated sarcopenia components, including handgrip strength, ASMi, gait speed, and SPPB score were significantly lower in participants with osteoporosis compared to participants with osteopenia or with normal bone status.

Table 2.

Sarcopenia components' comparison according to osteoporotic status

	n	Normal BMD and osteopenia	Osteoporosis*	p-value
		Results**
Handgrip (kg)	499	20.06 ± 4.9	17.6 ± 5.1	< 0.001
SPPB (/12 points)	499	10 (8–11)	9 (7–10)	< 0.001
Gait speed (m/s)	498	0.92 ± 0.25	0.83 ± 0.23	< 0.001
ASMi (kg/m²)	500	5.94 ± 0.91	5.60 ± 0.88	< 0.001

ASMi Appendicular Skeletal Muscle index, SPPB Short Physical Performances Battery

* According to the National Bone Health Alliance (reference 3)

** Mean ± SD or Medians with inter-quartile ranges, i.e. P50 (P25 - P75)

 Association between sarcopenia components and osteoporosis

Univariate analyses also showed significant associations between all sarcopenia components and osteoporotic status (Table 3). Handgrip strength below 16 kg was associated with a 2.5-fold increase in the odds of osteoporosis (95% CI, 1.59–3.94). Similarly, gait speed ≤ 0.8 m/s and ASMi < 5.5 kg/m² demonstrated significant associations with osteoporosis, with odds ratios of 2.10 (95% CI, 1.42–3.12) and 1.87 (95% CI, 1.26–2.77), respectively. Furthermore, an SPPB score ≤ 8 was associated with 1.67-fold greater odds of osteoporosis (95% CI, 1.13–2.49). In the multivariate analysis, including all these sarcopenia components, after adjustment for age, BMI, osteoporosis treatment, MHT, alcohol and tobacco intake, parental history of hip fracture, non-substituted early menopause, corticosteroid treatment and SARC-F score, significant associations were found between handgrip strength (p<0.01), gait speed (p<0.05), and the presence of osteoporosis. A handgrip strength < 16 kg was associated with 2.20-fold greater odds of osteoporosis (95% CI, 1.25–3.87), while gait speed ≤ 0.8 m/s was associated with 1.91-fold greater odds of osteoporosis (95% CI, 1.13–3.23). We also observed a significant association between sarcopenia defined by EWGSOP2 and osteoporosis with a crude odds ratio of 3.88 (95% CI, 1.97–7.85) (Table 4).

Table 3.

Logistic regression analyses for association of sarcopenia components (dichotomous) and osteoporosis* status

Sarcopenia components	n	Univariate analysis		Multivariate analysis**
		OR (95% CI)	p-value	OR (95% CI)	p-value
Handgrip strength < 16 kg	499	2.50 (1.59–3.94)	< 0.001	2.20 (1.25–3.87)	0.01
SPPB ≤ 8	499	1.67 (1.13–2.49)	0.01	1.15 (0.67–1.97)	0.60
Gait speed ≤ 0.8 m/s	498	2.10 (1.42–3.12)	< 0.001	1.91 (1.13–3.23)	0.02
ASMi < 5.5 kg/m²	496	1.87 (1.26–2.77)	0.001	0.89 (0.52–1.50)	0.65

ASMi Appendicular Skeletal Muscle index (ASM/height²), OR Odds-ratio, SPPB Short Physical Performances Battery

* According to the National Bone Health Alliance (reference 3)

**Adjusted for age, Body Mass Index, osteoporosis treatment, menopausal hormone therapy, alcohol and tobacco intake, parental history of hip fracture, non-substituted early menopause, corticosteroid therapy, and SARC-F score

Table 4.

Univariate analysis of the association between the definition of sarcopenia and osteoporosis as defined by the National Bone Health Alliance

	Osteoporosis*		p-value
	n (%)	OR (95% CI)
EWGSOP2
No sarcopenia	148 (32.6%)	Ref.	< 0.001
Sarcopenia	30 (65.2%)	3.88 (1.97–7.85)

EWGSOP2 Revised European consensus on definition and diagnosis of sarcopenia or European Working Group on Sarcopenia in Older People 2, OR Odds-ratio

*According to the National Bone Health Alliance (reference 3)

Discussion

Our study investigates the relationships between sarcopenia, its components, and osteoporosis as defined by the National Bone Health Alliance Working Group. The key findings were that sarcopenia, according to the EWGSOP2 criteria, and its components were significantly associated with osteoporosis, particularly handgrip strength and gait speed. We also observed significant correlations between the components of sarcopenia and BMD at all sites as well as with TBS.

In our cohort, composed of community-dwelling older women with a low level of frailty, the prevalence of sarcopenia was 9.2% based on the EWGSOP2 definition. This prevalence is higher than the one reported in a recent meta-analysis by Petermann-Rocha et al., which used the same diagnostic criteria. However, these authors also noted that the prevalence of sarcopenia can vary greatly, ranging from 0.3 to 91.2% in women, depending on the diagnostic criteria used. There is still no universally accepted definition of sarcopenia, leading to large differences in the results of various sarcopenia studies [21].

The operational definition of osteoporosis, based on the WHO criteria, includes participants with a BMD T-score ≤ −2.5 SD compared to the mean value for young adult women at the hip, femoral neck, or lumbar spine. In our study, despite ongoing debate about the recommended definition of osteoporosis, we decided to use the more inclusive definition of osteoporosis previously described by Siris et al., which includes individuals with a fragility MOF or a high predicted fracture risk [3]. Although this definition is not universally accepted, the American College of Endocrinology recently proposed similar diagnostic criteria to define osteoporosis in postmenopausal women [22]. As expected, the prevalence of osteoporosis in the study participants also varied depending on the criteria used; in our study, 35.6% of the participants had osteoporosis based on a BMD T-score ≤ −2.5, a history of MOF or high fracture risk, whereas only 23.6% would have been diagnosed with osteoporosis based solely on low BMD. This highlights and confirms the risk of underdiagnosis and undertreatment when BMD alone is considered as the diagnostic criterion.

Indeed, it is well-known that at least 50% of fragility fractures occur in patients who present a T-score greater than − 2.5 and even in those above − 1.0. Furthermore, BMD alone is insufficient to correctly predict fracture risk. Numerous other CRFs have been described, including age, corticosteroid treatment, parental history of hip fracture or a personal history of fragility fracture, which are notably used in the FRAX score or in the fracture prediction models recently developed by our team [9, 11, 23, 24]. In addition, pharmacological treatments, particularly bisphosphonates, have been proven to be useful in significantly reducing fracture risk and increasing BMD in patients with osteopenia without prior fractures [25]. This raises questions about the use of the standard osteoporosis definition and justifies our use of the clinical definition by the National Bone Health Alliance Working Group [3].

In our cohort study, a significant association between sarcopenia and its individually assessed components was observed with the presence of osteoporosis. After adjustment for confounding variables, only handgrip strength and gait speed remained significantly associated with osteoporosis. A collinear relationship was found between age, BMI, and sarcopenia components. Nevertheless, as in most previous studies, we decided to keep these covariates in multivariate models due to their significant relationship with BMD and clinical osteoporosis. In agreement with our results, other investigators have demonstrated a significant independent association between poor muscle strength, gait speed, and fracture risk [26–28]. Poor physical performance, defined by a low gait speed but also a low SPPB score has been shown to be associated with femoral neck and total hip BMD in a cohort of Chinese postmenopausal women [29].

Furthermore, low handgrip strength combined with slow gait speed has been associated with incident falls, mobility limitations, hip fractures, and death in a large American cohort study [30]. Handgrip strength has also been shown to predict fracture-free survival in women with a normal BMD in a 15-year follow-up study of Finnish women. Those women demonstrating handgrip strength results in the lowest quartile showed a lower fracture-free survival compared to the highest quartile [31]. All these observations suggest that handgrip strength and gait speed are strong predictors of increased fracture risk. Their easy assessment, availability, and inexpensiveness should make them useful and practical tools for detecting and preventing fractures in clinical practice in the older population.

We did not find a significant association between ASMi and osteoporosis after adjusting for various confounding covariates. The influence of body composition on osteoporosis and fracture risk is still debated due to inconsistent results. Various studies have reported similar findings to ours. In a cohort of 114 postmenopausal Brazilian women, ASMi did not remain significantly associated with BMD after adjustment for age, BMI, and race [32]. Kapus et al. also found no association between muscle mass or muscle strength and BMD, but they suggested an effect of fat mass on hip BMD [33]. In the subgroup of frail participants of the Women’s Health Initiative Observational Study, both muscle mass and fat mass were associated with hip BMD, but only higher fat mass was independently associated with a lower rate of hip fracture [34].

In contrast, other authors, such as Jang et al. [35] and Sjoblom et al. [36], reported that low muscle mass was significantly associated with osteoporosis in women, a finding consistent with the pathophysiological notion that bone anabolism is supported by the mechanical strain resulting from muscle contractions [37].

BMI could also interact with the association between muscle mass and BMD. Zhu et al. reported that muscle mass had a weaker association with BMD in the higher BMI tertile, particularly for BMD at the spine, suggesting an influence of fat mass, which is more closely correlated to BMI compared to muscle mass [38]. These inconsistent results could also be explained by the fact that DXA is not an ideal technique for measuring muscle mass, since it includes water and fibrotic tissue. The lack of association of DXA-measured muscle mass with adverse health outcomes may explain why it was not included in the Sarcopenia Definitions and Outcomes Consortium (SDOC) definition of sarcopenia, which only considers muscle slowness and weakness. Our results support this choice. However, muscle mass measured by more precise techniques such as D3-creatine dilution has demonstrated an association with significant health outcomes [30], leaving the question unresolved.

The association of criteria used to define sarcopenia with BMD and fracture risk has been extensively studied, but conflicting results persist in the literature. In a longitudinal study of almost 500 adults in The Netherlands, no association was found between sarcopenia, defined according to the FNIH or EWGSOP, or sarcopenia components, and fracture incidence. However, only 60 participants experienced a fracture during their follow-up, and the low statistical power to detect an association is probably an explanation for these results. Furthermore, contrary to our study, their cohort included 50% of men, though it is well known that women are more at risk of developing osteoporosis [39].

On the other hand, in agreement with our results, lower values in the components of sarcopenia were strongly associated with osteoporosis in the postmenopausal women of The Osteoporosis Risk Factor and Prevention - Fracture Prevention Study (OSTPRE-FPS) study [36]. A recent meta-analysis reported a significant association between sarcopenia and low BMD, as well as a high risk of fractures, which also supports the role of sarcopenia in assessing osteoporosis and emphasizes the importance of implementing sarcopenia prevention strategies alongside osteoporosis treatment [40]. Finally, in agreement with previous studies, we observed a significant correlation between sarcopenia components, specifically muscle strength and physical performance, and TBS. TBS is a textural analysis of lumbar spine DXA images reflecting bone microarchitecture and it is a risk factor for fracture, independent of BMD, in patients with osteopenia. An elevated TBS value indicates “stronger bones” and a lower risk of fracture. Similarly to our results, Locquet et al. reported a significant association between muscle strength, physical performance, and TBS in older women, although only muscle strength remained significantly associated after adjustment. Interestingly, Qi et al. demonstrated that handgrip strength showed a positive association with TBS. These findings suggest that these components are linked not only to bone mass but also to bone microarchitecture and strength [41, 42].

Five limitations should be acknowledged. First, the study’s cross-sectional nature makes it impossible to establish a causal relationship between sarcopenia or its components and osteoporosis. Second, adjusting covariates were selected based on statistical criteria: (i) their association with osteoporosis, as defined by the National Bone Health Alliance Working Group, and (ii) their impact on the association of EWGSOP2 sarcopenia and osteoporosis. Alternative model specifications, including the use of distinct adjustment variables for each sarcopenia component, may yield different results. Third, the study may be affected by biases, including survival bias, which could explain why the participants exhibit low frailty according to the Fried criteria. Our cohort mainly consisted of non-frail, community-dwelling older women willing to be included a second time in the FRISBEE study, limiting the generalizability of our findings to frail populations or to men. Additionally, potential reporting bias cannot be excluded, as self-reported alcohol intake may be underestimated. The small number of participants reporting alcohol consumption may limit the statistical power of this finding and alcohol consumers in our cohort were generally younger, which could reflect differences in social and health-related behaviors. Fourth, the FRAX score partly used to define clinical osteoporosis, was the score determined at baseline, which could underestimate the proportion of participants that suffer from osteoporosis in our cohort. Another limitation of the study was the lack of a comprehensive assessment of comorbidities that may influence musculoskeletal health, such as diabetes or kidney disease. Finally, while the study focuses on sarcopenia and osteoporosis with rigor, it does not address the recently described and emerging syndrome of osteosarcopenia, which may require further research [7].

The study also has three key strengths. First, the findings are highly innovative, as, to our knowledge, this is the first study to evaluate the association between sarcopenia and its components with a definition of osteoporosis extended to participants with a high fracture risk in the presence of normal or osteopenic BMD. Importantly, the study stands out due to its high methodological standards, both in study conduct, and analysis, as this manuscript is based on data from the FRISBEE cohort study, where all fractures, declared or not by the study participants, were validated by radiological reports [12]. Finally, the findings of the study align with recent literature suggesting a close link between bone and muscle, providing data that may help advance combined therapies for both sarcopenia and osteoporosis. Further research is needed to explore the relationships among osteosarcopenia and its components—both sarcopenia-related and osteoporosis-related—while considering the comprehensive range of existing consensual definitions for sarcopenia and osteoporosis, as well as its clinical associations with adverse outcomes.

Conclusions

We observed that sarcopenia and its components, defined by EWGSOP2, were significantly associated with osteoporosis, with handgrip strength and gait speed being strongly and independently associated with bone fragility. Handgrip strength and physical performance were also related to TBS and should therefore be considered when evaluating bone health and fracture risk. Our findings suggest that easily implementable and inexpensive tools are useful for detecting osteoporosis and could allow the early implementation of effective prevention and/or treatment strategies in clinical practice.

Further studies are needed to shed light on the crosstalk between bone and muscle, including specific data about osteoporosis, sarcopenia and its components, and incident fragility fractures, as one of the most clinically relevant outcomes in this population. Additionally, it remains to be established which measures of sarcopenia best predict fracture risk in order to be combined with classical CRFs and develop more precise prediction models for older and oldest-old populations.

Abbreviations

ASM

Appendicular Skeletal Muscle

ASMi

Appendicular Skeletal Muscle index

BMD

Bone mineral density

BMI

Body mass index

CIs

Confidence intervals

CRFs

Clinical risk factors

DXA

Dual-energy X-ray absorptiometry

IQR

Inter-quartile range

EWGSOP2

European Working Group on Sarcopenia in Older People 2

FNIH

Foundation for the National Institutes of Health

FRISBEE

Fracture RISk Brussels Epidemiological Enquiry cohort study

MHT

Menopausal hormone therapy

MMSE

Mini-mental state examination

MOF

Major osteoporotic fracture

OR

Odds ratio

OSTPRE-FPS

The Osteoporosis Risk Factor and Prevention - Fracture Prevention Study

PPI

Proton pump inhibitor

SD

Standard deviation

SDOC

Sarcopenia Definitions and Outcomes Consortium

SPPB

Short Physical Performance Battery

SSRI

Selective serotonin reuptake inhibitors

TBBM

Total-body bone mineral

TBS

Trabecular bone score

VIF

Variance inflation factors

WBF

Whole body fat

WHO

World Health Organization

Δd

Delta-deviance

Authors’ contributions

We confirm that all the authors participated in the preparation of the manuscript.TS, BJJ, BP, and BF designed the study. CA and BA did the statistical analyses. TS wrote the first draft of the manuscript. TS, BJJ, BP, and BF interpreted the findings. TS, BJJ, BP, FJ, IL, BA, SM, BF, ED, DSR, and HAS revised subsequent versions of the manuscript and contributed substantially. All authors read and approved the final version of the paper and have agreed to be accountable for the work.

Funding

The study was supported by grants from IRIS-Recherche, CHU Brugmann, Fondation Brugmann and Association Vésale.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request for common research purposes.

Declarations

Ethics approval and consent to participate

Written informed consent was obtained from all the participants. The study protocol was accepted by the Ethics Committee of all participating sites, Comité d’Ethique Hospitalier du C.H.U Brugmann (Number CE 2021/91).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.